agora inbox for pgsql-hackers@postgresql.org
help / color / mirror / Atom feed[PATCH v4 2/8] Address space reservation for shared memory
213+ messages / 2 participants
[nested] [flat]
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v4 2/8] Address space reservation for shared memory
@ 2024-10-16 18:21 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2024-10-16 18:21 UTC (permalink / raw)
Currently the kernel is responsible to chose an address, where to place each
shared memory mapping, which is the lowest possible address that do not clash
with any other mappings. This is considered to be the most portable approach,
but one of the downsides is that there is no place to resize allocated mappings
anymore. Here is how it looks like for one mapping in /proc/$PID/maps,
/dev/zero represents the anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
...
7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
By specifying the mapping address directly it's possible to place the
mapping in a way that leaves room for resizing. The idea is:
* To reserve some address space via mmap'ing a large chunk of memory
with PROT_NONE and MAP_NORESERVE. This way we prepare a playground for
preparing shared memory layout without risking anything interfering
with that.
* To slice the reserved space up into sections, one to use for each
shared segment.
* Allocate shared memory segments out of corresponding slices and
leaving unclaimed space in between them. This is implemented via
mmap'ing memory at a specified address from the reserved space with
MAP_FIXED.
The result looks like this:
012d9000-0133e000 [heap]
7f443a800000-7f444196c000 /dev/zero (deleted)
7f444196c000-7f470a800000 # reserved space
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
Things like address space randomization should not be a problem in this
context, since the randomization is applied to the mmap base, which is
one per process.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21250648 kB
VmRSS: 22948 kB
RssAnon: 768 kB
RssFile: 10404 kB
RssShmem: 11776 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
17637376 (~16.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
---
src/backend/port/sysv_shmem.c | 284 ++++++++++++++++++++++++----
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/storage/pg_shmem.h | 4 +-
6 files changed, 271 insertions(+), 39 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..a0f03ff868f 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -108,6 +108,66 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping placing (/dev/zero (deleted) below) looks like this:
+ *
+ * 00400000-00490000 /path/bin/postgres
+ * ...
+ * 012d9000-0133e000 [heap]
+ * 7f443a800000-7f470a800000 /dev/zero (deleted)
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * ...
+ * 7f471aef2000-7f471aef9000 /dev/shm/PostgreSQL.3859891842
+ * 7f471aef9000-7f471aefa000 /SYSV007dbf7d (deleted)
+ * ...
+ *
+ * We would like to place multiple mappings in such a way, that there will be
+ * enough space between them in the address space to be able to resize up to
+ * certain size, but without counting towards the total memory consumption.
+ *
+ * To achieve that we first reserve some shared memory address space by
+ * mmap'ing a segment of MaxAvailableMemory size with PROT_NONE and
+ * MAP_NORESERVE (these flags allow to make sure this space will not be used by
+ * anything else, yet do not count against memory limits). Having the reserved
+ * space, we allocate out of it actual chunks of shared memory as usual,
+ * updating a pointer to the current available reserved space for the next
+ * allocation with the gap between segments in mind.
+ *
+ * The result would look like this:
+ *
+ * 012d9000-0133e000 [heap]
+ * 7f4426f54000-7f442e010000 /dev/zero (deleted)
+ * 7f442e010000-7f443a800000 # reserved empty space
+ * 7f443a800000-7f444196c000 /dev/zero (deleted)
+ * 7f444196c000-7f470a800000 # reserved empty space
+ * 7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
+ * 7f4718400000-7f4718401000 /usr/lib64/libicudata.so.74.2
+ * [...]
+ *
+ * The reserved space pointer is calculated to slice up the total reserved
+ * space into fixed fractions of address space for each segment, as specified
+ * in the SHMEM_RESIZE_RATIO array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Offset from the beginning of the reserved space, which indicates currently
+ * available range. New shared memory segments have to be allocated at this
+ * offset related to the reserved space.
+ */
+static Size reserved_offset = 0;
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -626,39 +686,198 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
*
* This function will modify mapping size to the actual size of the allocation,
* if it ends up allocating a segment that is larger than requested.
+ *
+ * Note that we do not switch from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
-CreateAnonymousSegment(AnonymousMapping *mapping)
+CreateAnonymousSegment(AnonymousMapping *mapping, Pointer base)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
int mmap_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS;
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* ReserveAnonymousMemory should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
GetHugePageSize(&hugepagesize, &mmap_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
+ }
+#endif
+
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
+ /*
+ * Try to create mapping at an address out of the reserved range, which
+ * will allow to extend it later. Use reserved_offset to allocate the
+ * segment, then update currently available reserved range.
+ *
+ * If the last step has failed, fallback to the regular mapping
+ * creation and signal that shared buffers could not be resized without
+ * a restart.
+ */
+ ptr = mmap(base + reserved_offset, allocsize, PROT_READ | PROT_WRITE,
+ mmap_flags | MAP_FIXED, -1, 0);
+ mmap_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+ elog(DEBUG1, "segment[%s]: mmap(%zu) at address %p failed: %m, "
+ "fallback to the non-resizable allocation",
+ MappingName(mapping->shmem_segment), allocsize, base + reserved_offset);
+
ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
+ PG_MMAP_FLAGS, -1, 0);
+ mmap_errno = errno;
+ }
+ else
+ {
+ Size total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+
+ reserved_offset += total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+ }
+
+ if (ptr == MAP_FAILED)
+ {
+ errno = mmap_errno;
+ DebugMappings();
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment)),
+ (mmap_errno == ENOMEM) ?
+ errhint("This error usually means that PostgreSQL's request "
+ "for a shared memory segment exceeded available memory, "
+ "swap space, or huge pages. To reduce the request size "
+ "(currently %zu bytes), reduce PostgreSQL's shared "
+ "memory usage, perhaps by reducing \"shared_buffers\" or "
+ "\"max_connections\".",
+ allocsize) : 0));
+ }
+
+ mapping->shmem = ptr;
+ mapping->shmem_size = allocsize;
+}
+
+/*
+ * ReserveAnonymousMemory
+ *
+ * Reserve shared memory address space, from which shared memory segments are
+ * going to be sliced out. The goal of this exercise is to support segments
+ * resizing, for which we need a reserved space free of potential clashes with
+ * other mmap'd areas that are not under our control. Reservation is done via
+ * mmap, and will not allocate any memory until it will be actually used, and
+ * MAP_NORESERVE allows to make it not counting againt kernel reservation
+ * limits (e.g. in cgroups or for huge pages). Do not get confused because of
+ * MAP_NORESERVE -- we need to reserve some space, but not the actual memory,
+ * and that is that this flag is about.
+ *
+ * Note, that with MAP_NORESERVE a reservation with hugetlb will succeed even
+ * if there is actually not enough huge pages. Hence this function is
+ * responsible for deciding whether to use huge pages or not. To achieve that
+ * we need to probe first and try to allocate needed memory for all segments --
+ * if this succeeds, we unmap the probe segment and use hugetlb; if it fails,
+ * we proceed with the regular memory.
+ */
+void *
+ReserveAnonymousMemory(Size reserve_size)
+{
+ Size allocsize = reserve_size;
+ void *ptr = MAP_FAILED;
+ int mmap_errno = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ *
+ * We could actually have a mix and match of segments with and without
+ * huge pages. But in that case we need to have multiple reservation
+ * spaces to use corresponding memory (hugetlb adress space reserved
+ * for hugetlb segments, regular memory for others), and it doesn't
+ * seem to worth the complexity for now.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
mmap_errno = errno;
if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
{
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
+ /* No huge pages, we will go with the regular page size */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", total_size);
+ }
+ else
+ {
+ /*
+ * All fine, unmap the temporary segment and proceed with reserving
+ * using huge pages.
+ */
+ if (munmap(ptr, total_size) < 0)
+ elog(LOG, "reservice space: munmap(%p, %zu) failed: %m",
+ ptr, total_size);
+
+ /* Round up the requested size to a suitable large value. */
+ if (allocsize % hugepagesize != 0)
+ allocsize += hugepagesize - (allocsize % hugepagesize);
+
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB",
+ allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | MAP_NORESERVE | mmap_flags,
+ -1, 0);
+ mmap_errno = errno;
+
+ /* This should not happen, but handle errors anyway */
+ if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
+ {
+ elog(DEBUG1, "reserving space: mmap(%zu) with MAP_HUGETLB "
+ "failed, huge pages disabled: %m", allocsize);
+ }
}
}
#endif
@@ -666,10 +885,12 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
/*
* Report whether huge pages are in use. This needs to be tracked before
* the second mmap() call if attempting to use huge pages failed
- * previously.
+ * previously. At this point ptr is either pointing to the probe segment,
+ * if we couldn't mmap it, or the reservation space.
*/
SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
{
@@ -677,10 +898,11 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
* Use the original size, not the rounded-up value, when falling back
* to non-huge pages.
*/
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
+ allocsize = reserve_size;
+
+ elog(DEBUG1, "reserving space: mmap(%zu)", allocsize);
+ ptr = mmap(NULL, allocsize, PROT_NONE,
+ MAP_PRIVATE | MAP_ANONYMOUS | MAP_NORESERVE, -1, 0);
}
if (ptr == MAP_FAILED)
@@ -688,20 +910,18 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
errno = mmap_errno;
DebugMappings();
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
- MappingName(mapping->shmem_segment)),
+ (errmsg("reserving space: could not map anonymous shared "
+ "memory: %m"),
(mmap_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
- "for a shared memory segment exceeded available memory, "
- "swap space, or huge pages. To reduce the request size "
- "(currently %zu bytes), reduce PostgreSQL's shared "
- "memory usage, perhaps by reducing \"shared_buffers\" or "
- "\"max_connections\".",
+ "for a reserved shared memory address space exceeded "
+ "available memory, swap space, or huge pages. To "
+ "reduce the request reservation size (currently %zu "
+ "bytes), reduce PostgreSQL's \"maximum_shared_buffers\".",
allocsize) : 0));
}
- mapping->shmem = ptr;
- mapping->shmem_size = allocsize;
+ return ptr;
}
/*
@@ -740,7 +960,7 @@ AnonymousShmemDetach(int status, Datum arg)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
IpcMemoryKey NextShmemSegID;
void *memAddress;
@@ -760,14 +980,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -782,7 +994,7 @@ PGSharedMemoryCreate(Size size,
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
/* On success, mapping data will be modified. */
- CreateAnonymousSegment(mapping);
+ CreateAnonymousSegment(mapping, base);
next_free_segment++;
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..ce719f1b412 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -205,7 +205,7 @@ EnableLockPagesPrivilege(int elevel)
*/
PGShmemHeader *
PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim)
+ PGShmemHeader **shim, Pointer base)
{
void *memAddress;
PGShmemHeader *hdr;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..076888c0172 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -203,9 +203,12 @@ CreateSharedMemoryAndSemaphores(void)
PGShmemHeader *seghdr;
Size size;
int numSemas;
+ void *base;
Assert(!IsUnderPostmaster);
+ base = ReserveAnonymousMemory((Size) MaxAvailableMemory * BLCKSZ);
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -217,7 +220,7 @@ CreateSharedMemoryAndSemaphores(void)
*
* XXX: Do multiple shims are needed, one per segment?
*/
- seghdr = PGSharedMemoryCreate(size, &shim);
+ seghdr = PGSharedMemoryCreate(size, &shim, base);
/*
* Make sure that huge pages are never reported as "unknown" while the
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index 2152aad97d9..1d42a5856c0 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 131072;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index 4eaeca89f2c..dede37f7905 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2364,6 +2364,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 131072, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 138078c29c5..4a83e255652 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -60,6 +60,7 @@ extern PGDLLIMPORT ShmemSegment Segments[ANON_MAPPINGS];
extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -100,10 +101,11 @@ extern void PGSharedMemoryNoReAttach(void);
#endif
extern PGShmemHeader *PGSharedMemoryCreate(Size size,
- PGShmemHeader **shim);
+ PGShmemHeader **shim, Pointer base);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+void *ReserveAnonymousMemory(Size reserve_size);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.45.1
--vninua6xybvzgrci
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v4-0003-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH v5 05/10] Address space reservation for shared memory
@ 2025-06-17 09:47 Dmitrii Dolgov <9erthalion6@gmail.com>
0 siblings, 0 replies; 213+ messages in thread
From: Dmitrii Dolgov @ 2025-06-17 09:47 UTC (permalink / raw)
Currently the shared memory layout is designed to pack everything tight
together, leaving no space between mappings for resizing. Here is how it
looks like for one mapping in /proc/$PID/maps, /dev/zero represents the
anonymous shared memory we talk about:
00400000-00490000 /path/bin/postgres
...
012d9000-0133e000 [heap]
7f443a800000-7f470a800000 /dev/zero (deleted)
7f470a800000-7f471831d000 /usr/lib/locale/locale-archive
7f4718400000-7f4718401000 /usr/lib64/libstdc++.so.6.0.34
...
Make the layout more dynamic via splitting every shared memory segment
into two parts:
* An anonymous file, which actually contains shared memory content. Such
an anonymous file is created via memfd_create, it lives in memory,
behaves like a regular file and semantically equivalent to an
anonymous memory allocated via mmap with MAP_ANONYMOUS.
* A reservation mapping, which size is much larger than required shared
segment size. This mapping is created with flags PROT_NONE (which
makes sure the reserved space is not used), and MAP_NORESERVE (to not
count the reserved space against memory limits). The anonymous file is
mapped into this reservation mapping.
The resulting layout looks like this:
00400000-00490000 /path/bin/postgres
...
3f526000-3f590000 rw-p [heap]
7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted) -- anon file
7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted) -- reservation
7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
To resize a shared memory segment in this layout it's possible to use ftruncate
on the anonymous file, adjusting access permissions on the reserved space as
needed.
This approach also do not impact the actual memory usage as reported by
the kernel. Here is the output of /proc/$PID/status for the master
version with shared_buffers = 128 MB:
// Peak virtual memory size, which is described as total pages
// mapped in mm_struct. It corresponds to the mapped reserved space
// and is the only number that grows with it.
VmPeak: 2043192 kB
// Size of memory portions. It contains RssAnon + RssFile + RssShmem
VmRSS: 22908 kB
// Size of resident anonymous memory
RssAnon: 768 kB
// Size of resident file mappings
RssFile: 10364 kB
// Size of resident shmem memory (includes SysV shm, mapping of tmpfs and
// shared anonymous mappings)
RssShmem: 11776 kB
Here is the same for the patch when reserving 20GB of space:
VmPeak: 21255824 kB
VmRSS: 25020 kB
RssAnon: 768 kB
RssFile: 10812 kB
RssShmem: 13440 kB
Cgroup v2 doesn't have any problems with that as well. To verify a new cgroup
was created with the memory limit 256 MB, then PostgreSQL was launched withing
this cgroup with shared_buffers = 128 MB:
$ cd /sys/fs/cgroup
$ mkdir postgres
$ cd postres
$ echo 268435456 > memory.max
$ echo $MASTER_PID_SHELL > cgroup.procs
# postgres from the master branch has being successfully launched
# from that shell
$ cat memory.current
17465344 (~16.6 MB)
# stop postgres
$ echo $PATCH_PID_SHELL > cgroup.procs
# postgres from the patch has being successfully launched from that shell
$ cat memory.current
20770816 (~19.8 MB)
To control the amount of space reserved a new GUC max_available_memory
is introduced. Ideally it should be based on the maximum available
memory, hense the name.
There are also few unrelated advantages of using anon files:
* We've got a file descriptor, which could be used for regular file
operations (modification, truncation, you name it).
* The file could be given a name, which improves readability when it
comes to process maps.
* By default, Linux will not add file-backed shared mappings into a core dump,
making it more convenient to work with them in PostgreSQL: no more huge dumps
to process.
The downside is that memfd_create is Linux specific.
---
src/backend/port/sysv_shmem.c | 290 ++++++++++++++++++++++------
src/backend/port/win32_shmem.c | 2 +-
src/backend/storage/ipc/ipci.c | 5 +-
src/backend/storage/ipc/shmem.c | 2 +-
src/backend/utils/init/globals.c | 1 +
src/backend/utils/misc/guc_tables.c | 14 ++
src/include/miscadmin.h | 1 +
src/include/portability/mem.h | 2 +-
src/include/storage/pg_shmem.h | 5 +-
9 files changed, 262 insertions(+), 60 deletions(-)
diff --git a/src/backend/port/sysv_shmem.c b/src/backend/port/sysv_shmem.c
index 56af0231d24..363ddfd1fca 100644
--- a/src/backend/port/sysv_shmem.c
+++ b/src/backend/port/sysv_shmem.c
@@ -97,10 +97,12 @@ void *UsedShmemSegAddr = NULL;
typedef struct AnonymousMapping
{
int shmem_segment;
- Size shmem_size; /* Size of the mapping */
+ Size shmem_size; /* Size of the actually used memory */
+ Size shmem_reserved; /* Size of the reserved mapping */
Pointer shmem; /* Pointer to the start of the mapped memory */
Pointer seg_addr; /* SysV shared memory for the header */
unsigned long seg_id; /* IPC key */
+ int segment_fd; /* fd for the backing anon file */
} AnonymousMapping;
static AnonymousMapping Mappings[ANON_MAPPINGS];
@@ -108,6 +110,49 @@ static AnonymousMapping Mappings[ANON_MAPPINGS];
/* Keeps track of used mapping segments */
static int next_free_segment = 0;
+/*
+ * Anonymous mapping layout we use looks like this:
+ *
+ * 00400000-00c2a000 r-xp /bin/postgres
+ * ...
+ * 3f526000-3f590000 rw-p [heap]
+ * 7fbd827fe000-7fbd8bdde000 rw-s /memfd:main (deleted)
+ * 7fbd8bdde000-7fbe82800000 ---s /memfd:main (deleted)
+ * 7fbe82800000-7fbe90670000 r--p /usr/lib/locale/locale-archive
+ * 7fbe90800000-7fbe90941000 r-xp /usr/lib64/libstdc++.so.6.0.34
+ * ...
+ *
+ * We need to place shared memory mappings in such a way, that there will be
+ * gaps between them in the address space. Those gaps have to be large enough
+ * to resize the mapping up to certain size, without counting towards the total
+ * memory consumption.
+ *
+ * To achieve this, for each shared memory segment we first create an anonymous
+ * file of specified size using memfd_create, which will accomodate actual
+ * shared memory mapping content. It is represented by the first /memfd:main
+ * with rw permissions. Then we create a mapping for this file using mmap, with
+ * size much larger than required and flags PROT_NONE (allows to make sure the
+ * reserved space will not be used) and MAP_NORESERVE (prevents the space from
+ * being counted against memory limits). The mapping serves as an address space
+ * reservation, into which shared memory segment can be extended and is
+ * represented by the second /memfd:main with no permissions.
+ *
+ * The reserved space for each segment is calculated as a fraction of the total
+ * reserved space (MaxAvailableMemory), as specified in the SHMEM_RESIZE_RATIO
+ * array.
+ */
+static double SHMEM_RESIZE_RATIO[1] = {
+ 1.0, /* MAIN_SHMEM_SLOT */
+};
+
+/*
+ * Flag telling that we have decided to use huge pages.
+ *
+ * XXX: It's possible to use GetConfigOption("huge_pages_status", false, false)
+ * instead, but it feels like an overkill.
+ */
+static bool huge_pages_on = false;
+
static void *InternalIpcMemoryCreate(IpcMemoryKey memKey, Size size);
static void IpcMemoryDetach(int status, Datum shmaddr);
static void IpcMemoryDelete(int status, Datum shmId);
@@ -503,19 +548,20 @@ PGSharedMemoryAttach(IpcMemoryId shmId,
* hugepage sizes, we might want to think about more invasive strategies,
* such as increasing shared_buffers to absorb the extra space.
*
- * Returns the (real, assumed or config provided) page size into
- * *hugepagesize, and the hugepage-related mmap flags to use into
- * *mmap_flags if requested by the caller. If huge pages are not supported,
- * *hugepagesize and *mmap_flags are set to 0.
+ * Returns the (real, assumed or config provided) page size into *hugepagesize,
+ * the hugepage-related mmap and memfd flags to use into *mmap_flags and
+ * *memfd_flags if requested by the caller. If huge pages are not supported,
+ * *hugepagesize, *mmap_flags and *memfd_flags are set to 0.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
#ifdef MAP_HUGETLB
Size default_hugepagesize = 0;
Size hugepagesize_local = 0;
int mmap_flags_local = 0;
+ int memfd_flags_local = 0;
/*
* System-dependent code to find out the default huge page size.
@@ -574,6 +620,7 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
}
mmap_flags_local = MAP_HUGETLB;
+ memfd_flags_local = MFD_HUGETLB;
/*
* On recent enough Linux, also include the explicit page size, if
@@ -584,7 +631,16 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
{
int shift = pg_ceil_log2_64(hugepagesize_local);
- mmap_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
+ }
+#endif
+
+#if defined(MFD_HUGE_MASK) && defined(MFD_HUGE_SHIFT)
+ if (hugepagesize_local != default_hugepagesize)
+ {
+ int shift = pg_ceil_log2_64(hugepagesize_local);
+
+ memfd_flags_local |= (shift & MAP_HUGE_MASK) << MAP_HUGE_SHIFT;
}
#endif
@@ -593,6 +649,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*mmap_flags = mmap_flags_local;
if (hugepagesize)
*hugepagesize = hugepagesize_local;
+ if (memfd_flags)
+ *memfd_flags = memfd_flags_local;
#else
@@ -600,6 +658,8 @@ GetHugePageSize(Size *hugepagesize, int *mmap_flags)
*hugepagesize = 0;
if (mmap_flags)
*mmap_flags = 0;
+ if (memfd_flags)
+ *memfd_flags = 0;
#endif /* MAP_HUGETLB */
}
@@ -625,72 +685,90 @@ check_huge_page_size(int *newval, void **extra, GucSource source)
* Creates an anonymous mmap()ed shared memory segment.
*
* This function will modify mapping size to the actual size of the allocation,
- * if it ends up allocating a segment that is larger than requested.
+ * if it ends up allocating a segment that is larger than requested. If needed,
+ * it also rounds up the mapping reserved size to be a multiple of huge page
+ * size.
+ *
+ * Note that we do not fallback from huge pages to regular pages in this
+ * function, this decision was already made in ReserveAnonymousMemory and we
+ * stick to it.
*/
static void
CreateAnonymousSegment(AnonymousMapping *mapping)
{
Size allocsize = mapping->shmem_size;
void *ptr = MAP_FAILED;
- int mmap_errno = 0;
+ int save_errno = 0;
+ int mmap_flags = PG_MMAP_FLAGS, memfd_flags = 0;
+
+ elog(DEBUG1, "segment[%s]: size %zu, reserved %zu",
+ MappingName(mapping->shmem_segment), mapping->shmem_size,
+ mapping->shmem_reserved);
#ifndef MAP_HUGETLB
- /* PGSharedMemoryCreate should have dealt with this case */
- Assert(huge_pages != HUGE_PAGES_ON);
+ /* PrepareHugePages should have dealt with this case */
+ Assert(huge_pages != HUGE_PAGES_ON && !huge_pages_on);
#else
- if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ if (huge_pages_on)
{
- /*
- * Round up the request size to a suitable large value.
- */
Size hugepagesize;
- int mmap_flags;
- GetHugePageSize(&hugepagesize, &mmap_flags);
+ /* Make sure nothing is messed up */
+ Assert(huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY);
+
+ /* Round up the request size to a suitable large value */
+ GetHugePageSize(&hugepagesize, &mmap_flags, &memfd_flags);
if (allocsize % hugepagesize != 0)
allocsize += hugepagesize - (allocsize % hugepagesize);
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS | mmap_flags, -1, 0);
- mmap_errno = errno;
- if (huge_pages == HUGE_PAGES_TRY && ptr == MAP_FAILED)
- {
- DebugMappings();
- elog(DEBUG1, "segment[%s]: mmap(%zu) with MAP_HUGETLB failed, huge pages disabled: %m",
- MappingName(mapping->shmem_segment), allocsize);
- }
+ /*
+ * The reserved space is multiple of BLCKSZ. We know the huge page
+ * size, round up the reserved space to it.
+ */
+ mapping->shmem_reserved = mapping->shmem_reserved + hugepagesize -
+ (mapping->shmem_reserved % hugepagesize);
+
+ /* Verify that the new size is withing the reserved boundaries */
+ if (mapping->shmem_reserved < mapping->shmem_size)
+ ereport(ERROR,
+ (errcode(ERRCODE_INSUFFICIENT_RESOURCES),
+ errmsg("not enough shared memory is reserved"),
+ errhint("You may need to increase \"max_available_memory\".")));
+
+ mmap_flags = PG_MMAP_FLAGS | mmap_flags;
}
#endif
/*
- * Report whether huge pages are in use. This needs to be tracked before
- * the second mmap() call if attempting to use huge pages failed
- * previously.
+ * Prepare an anonymous file backing the segment. Its size will be
+ * specified later via ftruncate.
+ *
+ * The file behaves like a regular file, but lives in memory. Once all
+ * references to the file are dropped, it is automatically released.
+ * Anonymous memory is used for all backing pages of the file, thus it has
+ * the same semantics as anonymous memory allocations using mmap with the
+ * MAP_ANONYMOUS flag.
*/
- SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
- PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ mapping->segment_fd = memfd_create(MappingName(mapping->shmem_segment),
+ memfd_flags);
- if (ptr == MAP_FAILED && huge_pages != HUGE_PAGES_ON)
+ /*
+ * Specify the segment file size using allocsize, which contains
+ * potentially modified value.
+ */
+ if(ftruncate(mapping->segment_fd, allocsize) == -1)
{
- /*
- * Use the original size, not the rounded-up value, when falling back
- * to non-huge pages.
- */
- allocsize = mapping->shmem_size;
- ptr = mmap(NULL, allocsize, PROT_READ | PROT_WRITE,
- PG_MMAP_FLAGS, -1, 0);
- mmap_errno = errno;
- }
+ save_errno = errno;
- if (ptr == MAP_FAILED)
- {
- errno = mmap_errno;
DebugMappings();
+ close(mapping->segment_fd);
+
+ errno = save_errno;
ereport(FATAL,
- (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ (errmsg("segment[%s]: could not truncate anonymous file: %m",
MappingName(mapping->shmem_segment)),
- (mmap_errno == ENOMEM) ?
+ (save_errno == ENOMEM) ?
errhint("This error usually means that PostgreSQL's request "
"for a shared memory segment exceeded available memory, "
"swap space, or huge pages. To reduce the request size "
@@ -700,10 +778,112 @@ CreateAnonymousSegment(AnonymousMapping *mapping)
allocsize) : 0));
}
+ elog(DEBUG1, "segment[%s]: mmap(%zu)",
+ MappingName(mapping->shmem_segment), allocsize);
+
+ /*
+ * Create a reservation mapping.
+ */
+ ptr = mmap(NULL, mapping->shmem_reserved, PROT_NONE,
+ mmap_flags | MAP_NORESERVE, mapping->segment_fd, 0);
+ save_errno = errno;
+
+ if (ptr == MAP_FAILED)
+ {
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not map anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
+ /* Make the memory accessible */
+ if(mprotect(ptr, allocsize, PROT_READ | PROT_WRITE) == -1)
+ {
+ save_errno = errno;
+ DebugMappings();
+
+ errno = save_errno;
+ ereport(FATAL,
+ (errmsg("segment[%s]: could not mprotect anonymous shared memory: %m",
+ MappingName(mapping->shmem_segment))));
+ }
+
mapping->shmem = ptr;
mapping->shmem_size = allocsize;
}
+/*
+ * PrepareHugePages
+ *
+ * Figure out if there are enough huge pages to allocate all shared memory
+ * segments, and report that information via huge_pages_status and
+ * huge_pages_on. It needs to be called before creating shared memory segments.
+ *
+ * It is necessary to maintain the same semantic (simple on/off) for
+ * huge_pages_status, even if there are multiple shared memory segments: all
+ * segments either use huge pages or not, there is no mix of segments with
+ * different page size. The latter might be actually beneficial, in particular
+ * because only some segments may require large amount of memory, but for now
+ * we go with a simple solution.
+ */
+void
+PrepareHugePages()
+{
+ void *ptr = MAP_FAILED;
+
+ /* Reset to handle reinitialization */
+ next_free_segment = 0;
+
+ /* Complain if hugepages demanded but we can't possibly support them */
+#if !defined(MAP_HUGETLB)
+ if (huge_pages == HUGE_PAGES_ON)
+ ereport(ERROR,
+ (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
+ errmsg("huge pages not supported on this platform")));
+#else
+ if (huge_pages == HUGE_PAGES_ON || huge_pages == HUGE_PAGES_TRY)
+ {
+ Size hugepagesize, total_size = 0;
+ int mmap_flags;
+
+ GetHugePageSize(&hugepagesize, &mmap_flags, NULL);
+
+ /*
+ * Figure out how much memory is needed for all segments, keeping in
+ * mind that for every segment this value will be rounding up by the
+ * huge page size. The resulting value will be used to probe memory and
+ * decide whether we will allocate huge pages or not.
+ */
+ for(int segment = 0; segment < ANON_MAPPINGS; segment++)
+ {
+ int numSemas;
+ Size segment_size = CalculateShmemSize(&numSemas, segment);
+
+ if (segment_size % hugepagesize != 0)
+ segment_size += hugepagesize - (segment_size % hugepagesize);
+
+ total_size += segment_size;
+ }
+
+ /* Map total amount of memory to test its availability. */
+ elog(DEBUG1, "reserving space: probe mmap(%zu) with MAP_HUGETLB",
+ total_size);
+ ptr = mmap(NULL, total_size, PROT_NONE,
+ PG_MMAP_FLAGS | MAP_ANONYMOUS | mmap_flags, -1, 0);
+ }
+#endif
+
+ /*
+ * Report whether huge pages are in use. This needs to be tracked before
+ * creating shared memory segments.
+ */
+ SetConfigOption("huge_pages_status", (ptr == MAP_FAILED) ? "off" : "on",
+ PGC_INTERNAL, PGC_S_DYNAMIC_DEFAULT);
+ huge_pages_on = ptr != MAP_FAILED;
+}
+
/*
* AnonymousShmemDetach --- detach from an anonymous mmap'd block
* (called as an on_shmem_exit callback, hence funny argument list)
@@ -746,7 +926,7 @@ PGSharedMemoryCreate(Size size,
void *memAddress;
PGShmemHeader *hdr;
struct stat statbuf;
- Size sysvsize;
+ Size sysvsize, total_reserved;
AnonymousMapping *mapping = &Mappings[next_free_segment];
/*
@@ -760,14 +940,6 @@ PGSharedMemoryCreate(Size size,
errmsg("could not stat data directory \"%s\": %m",
DataDir)));
- /* Complain if hugepages demanded but we can't possibly support them */
-#if !defined(MAP_HUGETLB)
- if (huge_pages == HUGE_PAGES_ON)
- ereport(ERROR,
- (errcode(ERRCODE_FEATURE_NOT_SUPPORTED),
- errmsg("huge pages not supported on this platform")));
-#endif
-
/* For now, we don't support huge pages in SysV memory */
if (huge_pages == HUGE_PAGES_ON && shared_memory_type != SHMEM_TYPE_MMAP)
ereport(ERROR,
@@ -776,8 +948,16 @@ PGSharedMemoryCreate(Size size,
/* Room for a header? */
Assert(size > MAXALIGN(sizeof(PGShmemHeader)));
+
+ /* Prepare the mapping information */
mapping->shmem_size = size;
mapping->shmem_segment = next_free_segment;
+ total_reserved = (Size) MaxAvailableMemory * BLCKSZ;
+ mapping->shmem_reserved = total_reserved * SHMEM_RESIZE_RATIO[next_free_segment];
+
+ /* Round up to be a multiple of BLCKSZ */
+ mapping->shmem_reserved = mapping->shmem_reserved + BLCKSZ -
+ (mapping->shmem_reserved % BLCKSZ);
if (shared_memory_type == SHMEM_TYPE_MMAP)
{
diff --git a/src/backend/port/win32_shmem.c b/src/backend/port/win32_shmem.c
index 4dee856d6bd..732fedee87e 100644
--- a/src/backend/port/win32_shmem.c
+++ b/src/backend/port/win32_shmem.c
@@ -627,7 +627,7 @@ pgwin32_ReserveSharedMemoryRegion(HANDLE hChild)
* use GetLargePageMinimum() instead.
*/
void
-GetHugePageSize(Size *hugepagesize, int *mmap_flags)
+GetHugePageSize(Size *hugepagesize, int *mmap_flags, int *memfd_flags)
{
if (hugepagesize)
*hugepagesize = 0;
diff --git a/src/backend/storage/ipc/ipci.c b/src/backend/storage/ipc/ipci.c
index 8b38e985327..b60f7ef9ce2 100644
--- a/src/backend/storage/ipc/ipci.c
+++ b/src/backend/storage/ipc/ipci.c
@@ -206,6 +206,9 @@ CreateSharedMemoryAndSemaphores(void)
Assert(!IsUnderPostmaster);
+ /* Decide if we use huge pages or regular size pages */
+ PrepareHugePages();
+
for(int segment = 0; segment < ANON_MAPPINGS; segment++)
{
/* Compute the size of the shared-memory block */
@@ -377,7 +380,7 @@ InitializeShmemGUCs(void)
/*
* Calculate the number of huge pages required.
*/
- GetHugePageSize(&hp_size, NULL);
+ GetHugePageSize(&hp_size, NULL, NULL);
if (hp_size != 0)
{
Size hp_required;
diff --git a/src/backend/storage/ipc/shmem.c b/src/backend/storage/ipc/shmem.c
index 72255a1c5ca..8d025f0e907 100644
--- a/src/backend/storage/ipc/shmem.c
+++ b/src/backend/storage/ipc/shmem.c
@@ -817,7 +817,7 @@ pg_get_shmem_pagesize(void)
Assert(huge_pages_status != HUGE_PAGES_UNKNOWN);
if (huge_pages_status == HUGE_PAGES_ON)
- GetHugePageSize(&os_page_size, NULL);
+ GetHugePageSize(&os_page_size, NULL, NULL);
return os_page_size;
}
diff --git a/src/backend/utils/init/globals.c b/src/backend/utils/init/globals.c
index d31cb45a058..90d3feb547c 100644
--- a/src/backend/utils/init/globals.c
+++ b/src/backend/utils/init/globals.c
@@ -140,6 +140,7 @@ int max_parallel_maintenance_workers = 2;
* register background workers.
*/
int NBuffers = 16384;
+int MaxAvailableMemory = 524288;
int MaxConnections = 100;
int max_worker_processes = 8;
int max_parallel_workers = 8;
diff --git a/src/backend/utils/misc/guc_tables.c b/src/backend/utils/misc/guc_tables.c
index f04bfedb2fd..a221e446d6a 100644
--- a/src/backend/utils/misc/guc_tables.c
+++ b/src/backend/utils/misc/guc_tables.c
@@ -2376,6 +2376,20 @@ struct config_int ConfigureNamesInt[] =
NULL, NULL, NULL
},
+ {
+ {"max_available_memory", PGC_SIGHUP, RESOURCES_MEM,
+ gettext_noop("Sets the upper limit for the shared_buffers value."),
+ gettext_noop("Shared memory could be resized at runtime, this "
+ "parameters sets the upper limit for it, beyond which "
+ "resizing would not be supported. Normally this value "
+ "would be the same as the total available memory."),
+ GUC_UNIT_BLOCKS
+ },
+ &MaxAvailableMemory,
+ 524288, 16, INT_MAX / 2,
+ NULL, NULL, NULL
+ },
+
{
{"vacuum_buffer_usage_limit", PGC_USERSET, RESOURCES_MEM,
gettext_noop("Sets the buffer pool size for VACUUM, ANALYZE, and autovacuum."),
diff --git a/src/include/miscadmin.h b/src/include/miscadmin.h
index 1bef98471c3..a0c37a7749e 100644
--- a/src/include/miscadmin.h
+++ b/src/include/miscadmin.h
@@ -173,6 +173,7 @@ extern PGDLLIMPORT char *DataDir;
extern PGDLLIMPORT int data_directory_mode;
extern PGDLLIMPORT int NBuffers;
+extern PGDLLIMPORT int MaxAvailableMemory;
extern PGDLLIMPORT int MaxBackends;
extern PGDLLIMPORT int MaxConnections;
extern PGDLLIMPORT int max_worker_processes;
diff --git a/src/include/portability/mem.h b/src/include/portability/mem.h
index ef9800732d9..40588ff6968 100644
--- a/src/include/portability/mem.h
+++ b/src/include/portability/mem.h
@@ -38,7 +38,7 @@
#define MAP_NOSYNC 0
#endif
-#define PG_MMAP_FLAGS (MAP_SHARED|MAP_ANONYMOUS|MAP_HASSEMAPHORE)
+#define PG_MMAP_FLAGS (MAP_SHARED|MAP_HASSEMAPHORE)
/* Some really old systems don't define MAP_FAILED. */
#ifndef MAP_FAILED
diff --git a/src/include/storage/pg_shmem.h b/src/include/storage/pg_shmem.h
index 2348c59b5a0..79b0b1ef9eb 100644
--- a/src/include/storage/pg_shmem.h
+++ b/src/include/storage/pg_shmem.h
@@ -61,6 +61,7 @@ extern PGDLLIMPORT int shared_memory_type;
extern PGDLLIMPORT int huge_pages;
extern PGDLLIMPORT int huge_page_size;
extern PGDLLIMPORT int huge_pages_status;
+extern PGDLLIMPORT int MaxAvailableMemory;
/* Possible values for huge_pages and huge_pages_status */
typedef enum
@@ -104,7 +105,9 @@ extern PGShmemHeader *PGSharedMemoryCreate(Size size,
PGShmemHeader **shim);
extern bool PGSharedMemoryIsInUse(unsigned long id1, unsigned long id2);
extern void PGSharedMemoryDetach(void);
-extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags);
+extern void GetHugePageSize(Size *hugepagesize, int *mmap_flags,
+ int *memfd_flags);
+void PrepareHugePages(void);
/* The main segment, contains everything except buffer blocks and related data. */
#define MAIN_SHMEM_SEGMENT 0
--
2.49.0
--bwrlgp6w2ubxykjq
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment;
filename="v5-0006-Introduce-multiple-shmem-segments-for-shared-buff.patch"
^ permalink raw reply [nested|flat] 213+ messages in thread
* [PATCH 5/5] Use background worker to do logical decoding.
@ 2025-12-09 18:44 Antonin Houska <ah@cybertec.at>
0 siblings, 0 replies; 213+ messages in thread
From: Antonin Houska @ 2025-12-09 18:44 UTC (permalink / raw)
If the backend performing REPACK (CONCURRENTLY) does both data copying and
logical decoding, it has to "travel in time" back and forth and therefore it
has to invalidate system caches quite a few times. (The copying and the
decoding work with different catalog snapshots.) As the decoding worker has
separate caches, the switching is not necessary.
Without the worker, it'd also be difficult to switch between potentiallly long
running tasks like index build and WAL decoding. (No decoding during that time
at all can suspend archiving / recycling of WAL segments for some time, which
in turn may result in full disk.)
Another problem is that, after having acquired AccessExclusiveLock (in order
to swap the files), the backend needs to both decode and apply the data
changes that took place while it was waiting for the lock. With the decoding
worker, the decoding runs all the time, so the backend only needs to apply the
changes. This can reduce the time the exclusive lock is held for.
Note that the code added in order to handle ERRORs in the background worker
almost duplicates the existing code that does the same for other types of
workers (See ProcessParallelMessages() and
ProcessParallelApplyMessages()). Refactoring of the existing code might be
useful, to reduce the duplication.
---
src/backend/access/heap/heapam_handler.c | 44 -
src/backend/commands/cluster.c | 1139 +++++++++++++----
src/backend/libpq/pqmq.c | 5 +
src/backend/postmaster/bgworker.c | 4 +
src/backend/replication/logical/logical.c | 6 +-
.../pgoutput_repack/pgoutput_repack.c | 54 +-
src/backend/storage/ipc/procsignal.c | 4 +
src/backend/tcop/postgres.c | 4 +
.../utils/activity/wait_event_names.txt | 2 +
src/include/access/tableam.h | 7 +-
src/include/commands/cluster.h | 68 +-
src/include/storage/procsignal.h | 1 +
src/tools/pgindent/typedefs.list | 4 +-
13 files changed, 932 insertions(+), 410 deletions(-)
diff --git a/src/backend/access/heap/heapam_handler.c b/src/backend/access/heap/heapam_handler.c
index 01be29eb405..e6d630fa2f7 100644
--- a/src/backend/access/heap/heapam_handler.c
+++ b/src/backend/access/heap/heapam_handler.c
@@ -33,7 +33,6 @@
#include "catalog/index.h"
#include "catalog/storage.h"
#include "catalog/storage_xlog.h"
-#include "commands/cluster.h"
#include "commands/progress.h"
#include "executor/executor.h"
#include "miscadmin.h"
@@ -688,7 +687,6 @@ heapam_relation_copy_for_cluster(Relation OldHeap, Relation NewHeap,
Relation OldIndex, bool use_sort,
TransactionId OldestXmin,
Snapshot snapshot,
- LogicalDecodingContext *decoding_ctx,
TransactionId *xid_cutoff,
MultiXactId *multi_cutoff,
double *num_tuples,
@@ -710,7 +708,6 @@ heapam_relation_copy_for_cluster(Relation OldHeap, Relation NewHeap,
BufferHeapTupleTableSlot *hslot;
BlockNumber prev_cblock = InvalidBlockNumber;
bool concurrent = snapshot != NULL;
- XLogRecPtr end_of_wal_prev = GetFlushRecPtr(NULL);
/* Remember if it's a system catalog */
is_system_catalog = IsSystemRelation(OldHeap);
@@ -957,31 +954,6 @@ heapam_relation_copy_for_cluster(Relation OldHeap, Relation NewHeap,
ct_val[1] = *num_tuples;
pgstat_progress_update_multi_param(2, ct_index, ct_val);
}
-
- /*
- * Process the WAL produced by the load, as well as by other
- * transactions, so that the replication slot can advance and WAL does
- * not pile up. Use wal_segment_size as a threshold so that we do not
- * introduce the decoding overhead too often.
- *
- * Of course, we must not apply the changes until the initial load has
- * completed.
- *
- * Note that our insertions into the new table should not be decoded
- * as we (intentionally) do not write the logical decoding specific
- * information to WAL.
- */
- if (concurrent)
- {
- XLogRecPtr end_of_wal;
-
- end_of_wal = GetFlushRecPtr(NULL);
- if ((end_of_wal - end_of_wal_prev) > wal_segment_size)
- {
- repack_decode_concurrent_changes(decoding_ctx, end_of_wal);
- end_of_wal_prev = end_of_wal;
- }
- }
}
if (indexScan != NULL)
@@ -1027,22 +999,6 @@ heapam_relation_copy_for_cluster(Relation OldHeap, Relation NewHeap,
/* Report n_tuples */
pgstat_progress_update_param(PROGRESS_REPACK_HEAP_TUPLES_INSERTED,
n_tuples);
-
- /*
- * Try to keep the amount of not-yet-decoded WAL small, like
- * above.
- */
- if (concurrent)
- {
- XLogRecPtr end_of_wal;
-
- end_of_wal = GetFlushRecPtr(NULL);
- if ((end_of_wal - end_of_wal_prev) > wal_segment_size)
- {
- repack_decode_concurrent_changes(decoding_ctx, end_of_wal);
- end_of_wal_prev = end_of_wal;
- }
- }
}
tuplesort_end(tuplesort);
diff --git a/src/backend/commands/cluster.c b/src/backend/commands/cluster.c
index 501bd36c23e..f2a2ec6d3e5 100644
--- a/src/backend/commands/cluster.c
+++ b/src/backend/commands/cluster.c
@@ -46,6 +46,8 @@
#include "commands/tablecmds.h"
#include "commands/vacuum.h"
#include "executor/executor.h"
+#include "libpq/pqformat.h"
+#include "libpq/pqmq.h"
#include "miscadmin.h"
#include "optimizer/optimizer.h"
#include "pgstat.h"
@@ -56,6 +58,8 @@
#include "storage/ipc.h"
#include "storage/lmgr.h"
#include "storage/predicate.h"
+#include "storage/procsignal.h"
+#include "tcop/tcopprot.h"
#include "utils/acl.h"
#include "utils/fmgroids.h"
#include "utils/guc.h"
@@ -123,6 +127,103 @@ typedef struct ChangeDest
IndexInsertState *iistate;
} ChangeDest;
+/*
+ * Layout of shared memory used for communication between backend and the
+ * worker that performs logical decoding of data changes
+ */
+typedef struct DecodingWorkerShared
+{
+ /*
+ * Once the worker has reached this LSN, it should close the current
+ * output file and either create a new one or exit, according to the field
+ * 'done'. If the value is InvalidXLogRecPtr, the worker should decode all
+ * the WAL available and keep checking this field. It is ok if the worker
+ * had already decoded records whose LSN is >= lsn_upto before this field
+ * has been set.
+ */
+ XLogRecPtr lsn_upto;
+
+ /* Exit after closing the current file? */
+ bool done;
+
+ /* The output is stored here. */
+ SharedFileSet sfs;
+
+ /* Can backend read the file contents? */
+ bool sfs_valid;
+
+ /* Number of the last file exported by the worker. */
+ int last_exported;
+
+ /* Synchronize access to the fields above. */
+ slock_t mutex;
+
+ /* Database to connect to. */
+ Oid dbid;
+
+ /* Role to connect as. */
+ Oid roleid;
+
+ /* Decode data changes of this relation. */
+ Oid relid;
+
+ /* The backend uses this to wait for the worker. */
+ ConditionVariable cv;
+
+ /* Info to signal the backend. */
+ PGPROC *backend_proc;
+ pid_t backend_pid;
+ ProcNumber backend_proc_number;
+
+ /* Error queue. */
+ shm_mq *error_mq;
+
+ /*
+ * Memory the queue is located int.
+ *
+ * For considerations on the value see the comments of
+ * PARALLEL_ERROR_QUEUE_SIZE.
+ */
+#define REPACK_ERROR_QUEUE_SIZE 16384
+ char error_queue[FLEXIBLE_ARRAY_MEMBER];
+} DecodingWorkerShared;
+
+/*
+ * Generate output file name. If relations of the same 'relid' happen to be
+ * processed at the same time, they must be from different databases and
+ * therefore different backends must be involved. (PID is already present in
+ * the fileset name.)
+ */
+static inline void
+DecodingWorkerFileName(char *fname, Oid relid, uint32 seq)
+{
+ snprintf(fname, MAXPGPATH, "%u-%u", relid, seq);
+}
+
+/*
+ * Backend-local information to control the decoding worker.
+ */
+typedef struct DecodingWorker
+{
+ /* The worker. */
+ BackgroundWorkerHandle *handle;
+
+ /* DecodingWorkerShared is in this segment. */
+ dsm_segment *seg;
+
+ /* Handle of the error queue. */
+ shm_mq_handle *error_mqh;
+} DecodingWorker;
+
+/* Pointer to currently running decoding worker. */
+static DecodingWorker *decoding_worker = NULL;
+
+/*
+ * Is there a message sent by a repack worker that the backend needs to
+ * receive?
+ */
+volatile sig_atomic_t RepackMessagePending = false;
+
static bool cluster_rel_recheck(RepackCommand cmd, Relation OldHeap,
Oid indexOid, Oid userid, LOCKMODE lmode,
int options);
@@ -130,7 +231,7 @@ static void check_repack_concurrently_requirements(Relation rel);
static void rebuild_relation(Relation OldHeap, Relation index, bool verbose,
bool concurrent);
static void copy_table_data(Relation NewHeap, Relation OldHeap, Relation OldIndex,
- Snapshot snapshot, LogicalDecodingContext *decoding_ctx,
+ Snapshot snapshot,
bool verbose,
bool *pSwapToastByContent,
TransactionId *pFreezeXid,
@@ -143,12 +244,10 @@ static List *get_tables_to_repack_partitioned(RepackCommand cmd,
static bool cluster_is_permitted_for_relation(RepackCommand cmd,
Oid relid, Oid userid);
-static void begin_concurrent_repack(Relation rel);
-static void end_concurrent_repack(void);
static LogicalDecodingContext *setup_logical_decoding(Oid relid);
-static HeapTuple get_changed_tuple(char *change);
-static void apply_concurrent_changes(RepackDecodingState *dstate,
- ChangeDest *dest);
+static bool decode_concurrent_changes(LogicalDecodingContext *ctx,
+ DecodingWorkerShared *shared);
+static void apply_concurrent_changes(BufFile *file, ChangeDest *dest);
static void apply_concurrent_insert(Relation rel, HeapTuple tup,
IndexInsertState *iistate,
TupleTableSlot *index_slot);
@@ -160,9 +259,9 @@ static void apply_concurrent_delete(Relation rel, HeapTuple tup_target);
static HeapTuple find_target_tuple(Relation rel, ChangeDest *dest,
HeapTuple tup_key,
TupleTableSlot *ident_slot);
-static void process_concurrent_changes(LogicalDecodingContext *decoding_ctx,
- XLogRecPtr end_of_wal,
- ChangeDest *dest);
+static void process_concurrent_changes(XLogRecPtr end_of_wal,
+ ChangeDest *dest,
+ bool done);
static IndexInsertState *get_index_insert_state(Relation relation,
Oid ident_index_id,
Relation *ident_index_p);
@@ -172,7 +271,6 @@ static void free_index_insert_state(IndexInsertState *iistate);
static void cleanup_logical_decoding(LogicalDecodingContext *ctx);
static void rebuild_relation_finish_concurrent(Relation NewHeap, Relation OldHeap,
Relation cl_index,
- LogicalDecodingContext *decoding_ctx,
TransactionId frozenXid,
MultiXactId cutoffMulti);
static List *build_new_indexes(Relation NewHeap, Relation OldHeap, List *OldIndexes);
@@ -182,6 +280,13 @@ static Relation process_single_relation(RepackStmt *stmt,
ClusterParams *params);
static Oid determine_clustered_index(Relation rel, bool usingindex,
const char *indexname);
+static void start_decoding_worker(Oid relid);
+static void stop_decoding_worker(void);
+static void repack_worker_internal(dsm_segment *seg);
+static void export_initial_snapshot(Snapshot snapshot,
+ DecodingWorkerShared *shared);
+static Snapshot get_initial_snapshot(DecodingWorker *worker);
+static void ProcessRepackMessage(StringInfo msg);
static const char *RepackCommandAsString(RepackCommand cmd);
@@ -604,20 +709,20 @@ cluster_rel(RepackCommand cmd, Relation OldHeap, Oid indexOid,
/* rebuild_relation does all the dirty work */
PG_TRY();
{
- /*
- * For concurrent processing, make sure that our logical decoding
- * ignores data changes of other tables than the one we are
- * processing.
- */
- if (concurrent)
- begin_concurrent_repack(OldHeap);
-
rebuild_relation(OldHeap, index, verbose, concurrent);
}
PG_FINALLY();
{
if (concurrent)
- end_concurrent_repack();
+ {
+ /*
+ * Since during normal operation the worker was already asked to
+ * exit, stopping it explicitly is especially important on ERROR.
+ * However it still seems a good practice to make sure that the
+ * worker never survives the REPACK command.
+ */
+ stop_decoding_worker();
+ }
}
PG_END_TRY();
@@ -914,7 +1019,6 @@ rebuild_relation(Relation OldHeap, Relation index, bool verbose, bool concurrent
bool swap_toast_by_content;
TransactionId frozenXid;
MultiXactId cutoffMulti;
- LogicalDecodingContext *decoding_ctx = NULL;
Snapshot snapshot = NULL;
#if USE_ASSERT_CHECKING
LOCKMODE lmode;
@@ -928,19 +1032,36 @@ rebuild_relation(Relation OldHeap, Relation index, bool verbose, bool concurrent
if (concurrent)
{
/*
- * Prepare to capture the concurrent data changes.
+ * The worker needs to be member of the locking group we're the leader
+ * of. We ought to become the leader before the worker starts. The
+ * worker will join the group as soon as it starts.
+ *
+ * This is to make sure that the deadlock described below is
+ * detectable by deadlock.c: if the worker waits for a transaction to
+ * complete and we are waiting for the worker output, then effectively
+ * we (i.e. this backend) are waiting for that transaction.
+ */
+ BecomeLockGroupLeader();
+
+ /*
+ * Start the worker that decodes data changes applied while we're
+ * copying the table contents.
*
- * Note that this call waits for all transactions with XID already
- * assigned to finish. If some of those transactions is waiting for a
- * lock conflicting with ShareUpdateExclusiveLock on our table (e.g.
- * it runs CREATE INDEX), we can end up in a deadlock. Not sure this
- * risk is worth unlocking/locking the table (and its clustering
- * index) and checking again if its still eligible for REPACK
- * CONCURRENTLY.
+ * Note that the worker has to wait for all transactions with XID
+ * already assigned to finish. If some of those transactions is
+ * waiting for a lock conflicting with ShareUpdateExclusiveLock on our
+ * table (e.g. it runs CREATE INDEX), we can end up in a deadlock.
+ * Not sure this risk is worth unlocking/locking the table (and its
+ * clustering index) and checking again if its still eligible for
+ * REPACK CONCURRENTLY.
+ */
+ start_decoding_worker(tableOid);
+
+ /*
+ * Wait until the worker has the initial snapshot and retrieve it.
*/
- decoding_ctx = setup_logical_decoding(tableOid);
+ snapshot = get_initial_snapshot(decoding_worker);
- snapshot = SnapBuildInitialSnapshotForRepack(decoding_ctx->snapshot_builder);
PushActiveSnapshot(snapshot);
}
@@ -965,7 +1086,7 @@ rebuild_relation(Relation OldHeap, Relation index, bool verbose, bool concurrent
NewHeap = table_open(OIDNewHeap, NoLock);
/* Copy the heap data into the new table in the desired order */
- copy_table_data(NewHeap, OldHeap, index, snapshot, decoding_ctx, verbose,
+ copy_table_data(NewHeap, OldHeap, index, snapshot, verbose,
&swap_toast_by_content, &frozenXid, &cutoffMulti);
/* The historic snapshot won't be needed anymore. */
@@ -989,15 +1110,11 @@ rebuild_relation(Relation OldHeap, Relation index, bool verbose, bool concurrent
Assert(!swap_toast_by_content);
rebuild_relation_finish_concurrent(NewHeap, OldHeap, index,
- decoding_ctx,
frozenXid, cutoffMulti);
PopActiveSnapshot();
pgstat_progress_update_param(PROGRESS_REPACK_PHASE,
PROGRESS_REPACK_PHASE_FINAL_CLEANUP);
-
- /* Done with decoding. */
- cleanup_logical_decoding(decoding_ctx);
}
else
{
@@ -1168,8 +1285,7 @@ make_new_heap(Oid OIDOldHeap, Oid NewTableSpace, Oid NewAccessMethod,
*/
static void
copy_table_data(Relation NewHeap, Relation OldHeap, Relation OldIndex,
- Snapshot snapshot, LogicalDecodingContext *decoding_ctx,
- bool verbose, bool *pSwapToastByContent,
+ Snapshot snapshot, bool verbose, bool *pSwapToastByContent,
TransactionId *pFreezeXid, MultiXactId *pCutoffMulti)
{
Relation relRelation;
@@ -1330,7 +1446,6 @@ copy_table_data(Relation NewHeap, Relation OldHeap, Relation OldIndex,
*/
table_relation_copy_for_cluster(OldHeap, NewHeap, OldIndex, use_sort,
cutoffs.OldestXmin, snapshot,
- decoding_ctx,
&cutoffs.FreezeLimit,
&cutoffs.MultiXactCutoff,
&num_tuples, &tups_vacuumed,
@@ -2363,59 +2478,6 @@ RepackCommandAsString(RepackCommand cmd)
return "???";
}
-
-/*
- * Call this function before REPACK CONCURRENTLY starts to setup logical
- * decoding. It makes sure that other users of the table put enough
- * information into WAL.
- *
- * The point is that at various places we expect that the table we're
- * processing is treated like a system catalog. For example, we need to be
- * able to scan it using a "historic snapshot" anytime during the processing
- * (as opposed to scanning only at the start point of the decoding, as logical
- * replication does during initial table synchronization), in order to apply
- * concurrent UPDATE / DELETE commands.
- *
- * Note that TOAST table needs no attention here as it's not scanned using
- * historic snapshot.
- */
-static void
-begin_concurrent_repack(Relation rel)
-{
- Oid toastrelid;
-
- /*
- * Avoid logical decoding of other relations by this backend. The lock we
- * have guarantees that the actual locator cannot be changed concurrently:
- * TRUNCATE needs AccessExclusiveLock.
- */
- Assert(CheckRelationLockedByMe(rel, ShareUpdateExclusiveLock, false));
- repacked_rel_locator = rel->rd_locator;
- toastrelid = rel->rd_rel->reltoastrelid;
- if (OidIsValid(toastrelid))
- {
- Relation toastrel;
-
- /* Avoid logical decoding of other TOAST relations. */
- toastrel = table_open(toastrelid, AccessShareLock);
- repacked_rel_toast_locator = toastrel->rd_locator;
- table_close(toastrel, AccessShareLock);
- }
-}
-
-/*
- * Call this when done with REPACK CONCURRENTLY.
- */
-static void
-end_concurrent_repack(void)
-{
- /*
- * Restore normal function of (future) logical decoding for this backend.
- */
- repacked_rel_locator.relNumber = InvalidOid;
- repacked_rel_toast_locator.relNumber = InvalidOid;
-}
-
/*
* This function is much like pg_create_logical_replication_slot() except that
* the new slot is neither released (if anyone else could read changes from
@@ -2427,9 +2489,10 @@ static LogicalDecodingContext *
setup_logical_decoding(Oid relid)
{
Relation rel;
- TupleDesc tupdesc;
+ Oid toastrelid;
LogicalDecodingContext *ctx;
- RepackDecodingState *dstate = palloc0_object(RepackDecodingState);
+ NameData slotname;
+ RepackDecodingState *dstate;
/*
* REPACK CONCURRENTLY is not allowed in a transaction block, so this
@@ -2437,21 +2500,21 @@ setup_logical_decoding(Oid relid)
*/
Assert(!TransactionIdIsValid(GetTopTransactionIdIfAny()));
- /*
- * A single backend should not execute multiple REPACK commands at a time,
- * so use PID to make the slot unique.
- */
- snprintf(NameStr(dstate->slotname), NAMEDATALEN, "repack_%d", MyProcPid);
-
/*
* Check if we can use logical decoding.
*/
CheckSlotPermissions();
CheckLogicalDecodingRequirements();
- /* RS_TEMPORARY so that the slot gets cleaned up on ERROR. */
- ReplicationSlotCreate(NameStr(dstate->slotname), true, RS_TEMPORARY,
- false, false, false);
+ /*
+ * A single backend should not execute multiple REPACK commands at a time,
+ * so use PID to make the slot unique.
+ *
+ * RS_TEMPORARY so that the slot gets cleaned up on ERROR.
+ */
+ snprintf(NameStr(slotname), NAMEDATALEN, "repack_%d", MyProcPid);
+ ReplicationSlotCreate(NameStr(slotname), true, RS_TEMPORARY, false, false,
+ false);
/*
* Neither prepare_write nor do_write callback nor update_progress is
@@ -2473,197 +2536,238 @@ setup_logical_decoding(Oid relid)
DecodingContextFindStartpoint(ctx);
+ /*
+ * decode_concurrent_changes() needs non-blocking callback.
+ */
+ ctx->reader->routine.page_read = read_local_xlog_page_no_wait;
+
+ /*
+ * read_local_xlog_page_no_wait() needs to be able to indicate the end of
+ * WAL.
+ */
+ ctx->reader->private_data = MemoryContextAllocZero(ctx->context,
+ sizeof(ReadLocalXLogPageNoWaitPrivate));
+
+
/* Some WAL records should have been read. */
Assert(ctx->reader->EndRecPtr != InvalidXLogRecPtr);
+ /*
+ * Initialize repack_current_segment so that we can notice WAL segment
+ * boundaries.
+ */
XLByteToSeg(ctx->reader->EndRecPtr, repack_current_segment,
wal_segment_size);
- /*
- * Setup structures to store decoded changes.
- */
+ dstate = palloc0_object(RepackDecodingState);
dstate->relid = relid;
- dstate->tstore = tuplestore_begin_heap(false, false,
- maintenance_work_mem);
- /* Caller should already have the table locked. */
- rel = table_open(relid, NoLock);
- tupdesc = CreateTupleDescCopy(RelationGetDescr(rel));
- dstate->tupdesc = tupdesc;
- table_close(rel, NoLock);
+ /*
+ * Tuple descriptor may be needed to flatten a tuple before we write it to
+ * a file. A copy is needed because the decoding worker invalidates system
+ * caches before it starts to do the actual work.
+ */
+ rel = table_open(relid, AccessShareLock);
+ dstate->tupdesc = CreateTupleDescCopy(RelationGetDescr(rel));
- /* Initialize the descriptor to store the changes ... */
- dstate->tupdesc_change = CreateTemplateTupleDesc(1);
+ /* Avoid logical decoding of other relations. */
+ repacked_rel_locator = rel->rd_locator;
+ toastrelid = rel->rd_rel->reltoastrelid;
+ if (OidIsValid(toastrelid))
+ {
+ Relation toastrel;
- TupleDescInitEntry(dstate->tupdesc_change, 1, NULL, BYTEAOID, -1, 0);
- /* ... as well as the corresponding slot. */
- dstate->tsslot = MakeSingleTupleTableSlot(dstate->tupdesc_change,
- &TTSOpsMinimalTuple);
+ /* Avoid logical decoding of other TOAST relations. */
+ toastrel = table_open(toastrelid, AccessShareLock);
+ repacked_rel_toast_locator = toastrel->rd_locator;
+ table_close(toastrel, AccessShareLock);
+ }
+ table_close(rel, AccessShareLock);
- dstate->resowner = ResourceOwnerCreate(CurrentResourceOwner,
- "logical decoding");
+ /* The file will be set as soon as we have it opened. */
+ dstate->file = NULL;
ctx->output_writer_private = dstate;
+
return ctx;
}
/*
- * Retrieve tuple from ConcurrentChange structure.
+ * Decode logical changes from the WAL sequence and store them to a file.
*
- * The input data starts with the structure but it might not be appropriately
- * aligned.
+ * If true is returned, there is no more work for the worker.
*/
-static HeapTuple
-get_changed_tuple(char *change)
+static bool
+decode_concurrent_changes(LogicalDecodingContext *ctx,
+ DecodingWorkerShared *shared)
{
- HeapTupleData tup_data;
- HeapTuple result;
- char *src;
+ RepackDecodingState *dstate;
+ XLogRecPtr lsn_upto;
+ bool done;
+ char fname[MAXPGPATH];
- /*
- * Ensure alignment before accessing the fields. (This is why we can't use
- * heap_copytuple() instead of this function.)
- */
- src = change + offsetof(ConcurrentChange, tup_data);
- memcpy(&tup_data, src, sizeof(HeapTupleData));
+ dstate = (RepackDecodingState *) ctx->output_writer_private;
- result = (HeapTuple) palloc(HEAPTUPLESIZE + tup_data.t_len);
- memcpy(result, &tup_data, sizeof(HeapTupleData));
- result->t_data = (HeapTupleHeader) ((char *) result + HEAPTUPLESIZE);
- src = change + SizeOfConcurrentChange;
- memcpy(result->t_data, src, result->t_len);
+ /* Open the output file. */
+ DecodingWorkerFileName(fname, shared->relid, shared->last_exported + 1);
+ dstate->file = BufFileCreateFileSet(&shared->sfs.fs, fname);
- return result;
-}
+ SpinLockAcquire(&shared->mutex);
+ lsn_upto = shared->lsn_upto;
+ done = shared->done;
+ SpinLockRelease(&shared->mutex);
-/*
- * Decode logical changes from the WAL sequence up to end_of_wal.
- */
-void
-repack_decode_concurrent_changes(LogicalDecodingContext *ctx,
- XLogRecPtr end_of_wal)
-{
- RepackDecodingState *dstate;
- ResourceOwner resowner_old;
-
- /*
- * Invalidate the "present" cache before moving to "(recent) history".
- */
- InvalidateSystemCaches();
+ while (XLogRecPtrIsInvalid(lsn_upto) || ctx->reader->EndRecPtr < lsn_upto)
+ {
+ XLogRecord *record;
+ XLogSegNo segno_new;
+ char *errm = NULL;
+ XLogRecPtr end_lsn;
- dstate = (RepackDecodingState *) ctx->output_writer_private;
- resowner_old = CurrentResourceOwner;
- CurrentResourceOwner = dstate->resowner;
+ CHECK_FOR_INTERRUPTS();
- PG_TRY();
- {
- while (ctx->reader->EndRecPtr < end_of_wal)
+ record = XLogReadRecord(ctx->reader, &errm);
+ if (record == NULL)
{
- XLogRecord *record;
- XLogSegNo segno_new;
- char *errm = NULL;
- XLogRecPtr end_lsn;
+ ReadLocalXLogPageNoWaitPrivate *priv;
- record = XLogReadRecord(ctx->reader, &errm);
if (errm)
- elog(ERROR, "%s", errm);
-
- if (record != NULL)
- LogicalDecodingProcessRecord(ctx, ctx->reader);
+ ereport(ERROR, (errmsg("%s", errm)));
/*
- * If WAL segment boundary has been crossed, inform the decoding
- * system that the catalog_xmin can advance. (We can confirm more
- * often, but a filling a single WAL segment should not take much
- * time.)
+ * In the decoding loop we do not want to get blocked when there
+ * is no more WAL available, otherwise the loop would become
+ * uninterruptible. The point is that the worker is only useful if
+ * it starts decoding before lsn_upto is set. Thus it can reach
+ * the end of WAL and find out later that it did not have to go
+ * that far.
*/
- end_lsn = ctx->reader->EndRecPtr;
- XLByteToSeg(end_lsn, segno_new, wal_segment_size);
- if (segno_new != repack_current_segment)
+ priv = (ReadLocalXLogPageNoWaitPrivate *)
+ ctx->reader->private_data;
+ if (priv->end_of_wal)
{
- LogicalConfirmReceivedLocation(end_lsn);
- elog(DEBUG1, "REPACK: confirmed receive location %X/%X",
- (uint32) (end_lsn >> 32), (uint32) end_lsn);
- repack_current_segment = segno_new;
+ priv->end_of_wal = false;
+
+ /* Do we know how far we should get? */
+ if (XLogRecPtrIsInvalid(lsn_upto))
+ {
+ SpinLockAcquire(&shared->mutex);
+ lsn_upto = shared->lsn_upto;
+ /* 'done' should be set at the same time as 'lsn_upto' */
+ done = shared->done;
+ SpinLockRelease(&shared->mutex);
+
+ /* Check if the work happens to be complete. */
+ continue;
+ }
+
+ /* Wait a bit before we retry reading WAL. */
+ (void) WaitLatch(MyLatch,
+ WL_LATCH_SET | WL_TIMEOUT | WL_EXIT_ON_PM_DEATH,
+ 1000L,
+ WAIT_EVENT_REPACK_WORKER_MAIN);
+
+ continue;
}
+ else
+ ereport(ERROR, (errmsg("could not read WAL record")));
+ }
- CHECK_FOR_INTERRUPTS();
+ LogicalDecodingProcessRecord(ctx, ctx->reader);
+
+ /*
+ * If WAL segment boundary has been crossed, inform the decoding
+ * system that the catalog_xmin can advance.
+ *
+ * TODO Does it make sense to confirm more often? Segment size seems
+ * appropriate for restart_lsn (because less than a segment cannot be
+ * recycled anyway), however more frequent checks might be beneficial
+ * for catalog_xmin.
+ */
+ end_lsn = ctx->reader->EndRecPtr;
+ XLByteToSeg(end_lsn, segno_new, wal_segment_size);
+ if (segno_new != repack_current_segment)
+ {
+ LogicalConfirmReceivedLocation(end_lsn);
+ elog(DEBUG1, "REPACK: confirmed receive location %X/%X",
+ (uint32) (end_lsn >> 32), (uint32) end_lsn);
+ repack_current_segment = segno_new;
+ }
+
+ /* Keep checking if 'lsn_upto' was specified. */
+ if (XLogRecPtrIsInvalid(lsn_upto))
+ {
+ SpinLockAcquire(&shared->mutex);
+ lsn_upto = shared->lsn_upto;
+ /* 'done' should be set at the same time as 'lsn_upto' */
+ done = shared->done;
+ SpinLockRelease(&shared->mutex);
}
- InvalidateSystemCaches();
- CurrentResourceOwner = resowner_old;
- }
- PG_CATCH();
- {
- /* clear all timetravel entries */
- InvalidateSystemCaches();
- CurrentResourceOwner = resowner_old;
- PG_RE_THROW();
}
- PG_END_TRY();
+
+ /*
+ * Close the file and make it available to the backend.
+ */
+ BufFileClose(dstate->file);
+ dstate->file = NULL;
+ SpinLockAcquire(&shared->mutex);
+ shared->lsn_upto = InvalidXLogRecPtr;
+ shared->sfs_valid = true;
+ shared->last_exported++;
+ SpinLockRelease(&shared->mutex);
+ ConditionVariableSignal(&shared->cv);
+
+ return done;
}
/*
- * Apply changes that happened during the initial load.
- *
- * Scan key is passed by caller, so it does not have to be constructed
- * multiple times. Key entries have all fields initialized, except for
- * sk_argument.
+ * Apply changes stored in 'file'.
*/
static void
-apply_concurrent_changes(RepackDecodingState *dstate, ChangeDest *dest)
+apply_concurrent_changes(BufFile *file, ChangeDest *dest)
{
+ char kind;
+ uint32 t_len;
Relation rel = dest->rel;
TupleTableSlot *index_slot,
*ident_slot;
HeapTuple tup_old = NULL;
- if (dstate->nchanges == 0)
- return;
-
/* TupleTableSlot is needed to pass the tuple to ExecInsertIndexTuples(). */
- index_slot = MakeSingleTupleTableSlot(dstate->tupdesc, &TTSOpsHeapTuple);
+ index_slot = MakeSingleTupleTableSlot(RelationGetDescr(rel),
+ &TTSOpsHeapTuple);
/* A slot to fetch tuples from identity index. */
ident_slot = table_slot_create(rel, NULL);
- while (tuplestore_gettupleslot(dstate->tstore, true, false,
- dstate->tsslot))
+ while (true)
{
- bool shouldFree;
- HeapTuple tup_change,
- tup,
+ size_t nread;
+ HeapTuple tup,
tup_exist;
- char *change_raw,
- *src;
- ConcurrentChange change;
- bool isnull[1];
- Datum values[1];
CHECK_FOR_INTERRUPTS();
- /* Get the change from the single-column tuple. */
- tup_change = ExecFetchSlotHeapTuple(dstate->tsslot, false, &shouldFree);
- heap_deform_tuple(tup_change, dstate->tupdesc_change, values, isnull);
- Assert(!isnull[0]);
-
- /* Make sure we access aligned data. */
- change_raw = (char *) DatumGetByteaP(values[0]);
- src = (char *) VARDATA(change_raw);
- memcpy(&change, src, SizeOfConcurrentChange);
+ nread = BufFileReadMaybeEOF(file, &kind, 1, true);
+ /* Are we done with the file? */
+ if (nread == 0)
+ break;
- /*
- * Extract the tuple from the change. The tuple is copied here because
- * it might be assigned to 'tup_old', in which case it needs to
- * survive into the next iteration.
- */
- tup = get_changed_tuple(src);
+ /* Read the tuple. */
+ BufFileReadExact(file, &t_len, sizeof(t_len));
+ tup = (HeapTuple) palloc(HEAPTUPLESIZE + t_len);
+ tup->t_data = (HeapTupleHeader) ((char *) tup + HEAPTUPLESIZE);
+ BufFileReadExact(file, tup->t_data, t_len);
+ tup->t_len = t_len;
+ ItemPointerSetInvalid(&tup->t_self);
+ tup->t_tableOid = RelationGetRelid(dest->rel);
- if (change.kind == CHANGE_UPDATE_OLD)
+ if (kind == CHANGE_UPDATE_OLD)
{
Assert(tup_old == NULL);
tup_old = tup;
}
- else if (change.kind == CHANGE_INSERT)
+ else if (kind == CHANGE_INSERT)
{
Assert(tup_old == NULL);
@@ -2671,12 +2775,11 @@ apply_concurrent_changes(RepackDecodingState *dstate, ChangeDest *dest)
pfree(tup);
}
- else if (change.kind == CHANGE_UPDATE_NEW ||
- change.kind == CHANGE_DELETE)
+ else if (kind == CHANGE_UPDATE_NEW || kind == CHANGE_DELETE)
{
HeapTuple tup_key;
- if (change.kind == CHANGE_UPDATE_NEW)
+ if (kind == CHANGE_UPDATE_NEW)
{
tup_key = tup_old != NULL ? tup_old : tup;
}
@@ -2693,7 +2796,7 @@ apply_concurrent_changes(RepackDecodingState *dstate, ChangeDest *dest)
if (tup_exist == NULL)
elog(ERROR, "failed to find target tuple");
- if (change.kind == CHANGE_UPDATE_NEW)
+ if (kind == CHANGE_UPDATE_NEW)
apply_concurrent_update(rel, tup, tup_exist, dest->iistate,
index_slot);
else
@@ -2708,26 +2811,19 @@ apply_concurrent_changes(RepackDecodingState *dstate, ChangeDest *dest)
pfree(tup);
}
else
- elog(ERROR, "unrecognized kind of change: %d", change.kind);
+ elog(ERROR, "unrecognized kind of change: %d", kind);
/*
* If a change was applied now, increment CID for next writes and
* update the snapshot so it sees the changes we've applied so far.
*/
- if (change.kind != CHANGE_UPDATE_OLD)
+ if (kind != CHANGE_UPDATE_OLD)
{
CommandCounterIncrement();
UpdateActiveSnapshotCommandId();
}
-
- /* TTSOpsMinimalTuple has .get_heap_tuple==NULL. */
- Assert(shouldFree);
- pfree(tup_change);
}
- tuplestore_clear(dstate->tstore);
- dstate->nchanges = 0;
-
/* Cleanup. */
ExecDropSingleTupleTableSlot(index_slot);
ExecDropSingleTupleTableSlot(ident_slot);
@@ -2865,6 +2961,12 @@ find_target_tuple(Relation rel, ChangeDest *dest, HeapTuple tup_key,
/* XXX no instrumentation for now */
scan = index_beginscan(rel, ident_index, GetActiveSnapshot(),
NULL, dest->ident_key_nentries, 0);
+
+ /*
+ * Scan key is passed by caller, so it does not have to be constructed
+ * multiple times. Key entries have all fields initialized, except for
+ * sk_argument.
+ */
index_rescan(scan, dest->ident_key, dest->ident_key_nentries, NULL, 0);
/* Info needed to retrieve key values from heap tuple. */
@@ -2900,25 +3002,58 @@ find_target_tuple(Relation rel, ChangeDest *dest, HeapTuple tup_key,
}
/*
- * Decode and apply concurrent changes.
+ * Decode and apply concurrent changes, up to (and including) the record whose
+ * LSN is 'end_of_wal'.
*/
static void
-process_concurrent_changes(LogicalDecodingContext *decoding_ctx,
- XLogRecPtr end_of_wal, ChangeDest *dest)
+process_concurrent_changes(XLogRecPtr end_of_wal, ChangeDest *dest, bool done)
{
- RepackDecodingState *dstate;
+ DecodingWorkerShared *shared;
+ char fname[MAXPGPATH];
+ BufFile *file;
pgstat_progress_update_param(PROGRESS_REPACK_PHASE,
PROGRESS_REPACK_PHASE_CATCH_UP);
- dstate = (RepackDecodingState *) decoding_ctx->output_writer_private;
+ /* Ask the worker for the file. */
+ shared = (DecodingWorkerShared *) dsm_segment_address(decoding_worker->seg);
+ SpinLockAcquire(&shared->mutex);
+ shared->lsn_upto = end_of_wal;
+ shared->done = done;
+ SpinLockRelease(&shared->mutex);
- repack_decode_concurrent_changes(decoding_ctx, end_of_wal);
+ /*
+ * The worker needs to finish processing of the current WAL record. Even
+ * if it's idle, it'll need to close the output file. Thus we're likely to
+ * wait, so prepare for sleep.
+ */
+ ConditionVariablePrepareToSleep(&shared->cv);
+ for (;;)
+ {
+ bool valid;
- if (dstate->nchanges == 0)
- return;
+ SpinLockAcquire(&shared->mutex);
+ valid = shared->sfs_valid;
+ SpinLockRelease(&shared->mutex);
+
+ if (valid)
+ break;
+
+ ConditionVariableSleep(&shared->cv, WAIT_EVENT_REPACK_WORKER_EXPORT);
+ }
+ ConditionVariableCancelSleep();
- apply_concurrent_changes(dstate, dest);
+ /* Open the file. */
+ DecodingWorkerFileName(fname, shared->relid, shared->last_exported);
+ file = BufFileOpenFileSet(&shared->sfs.fs, fname, O_RDONLY, false);
+ apply_concurrent_changes(file, dest);
+
+ /* No file is exported until the worker exports the next one. */
+ SpinLockAcquire(&shared->mutex);
+ shared->sfs_valid = false;
+ SpinLockRelease(&shared->mutex);
+
+ BufFileClose(file);
}
/*
@@ -3044,15 +3179,10 @@ cleanup_logical_decoding(LogicalDecodingContext *ctx)
dstate = (RepackDecodingState *) ctx->output_writer_private;
- ExecDropSingleTupleTableSlot(dstate->tsslot);
- FreeTupleDesc(dstate->tupdesc_change);
FreeTupleDesc(dstate->tupdesc);
- tuplestore_end(dstate->tstore);
-
FreeDecodingContext(ctx);
- ReplicationSlotRelease();
- ReplicationSlotDrop(NameStr(dstate->slotname), false);
+ ReplicationSlotDropAcquired();
pfree(dstate);
}
@@ -3067,7 +3197,6 @@ cleanup_logical_decoding(LogicalDecodingContext *ctx)
static void
rebuild_relation_finish_concurrent(Relation NewHeap, Relation OldHeap,
Relation cl_index,
- LogicalDecodingContext *decoding_ctx,
TransactionId frozenXid,
MultiXactId cutoffMulti)
{
@@ -3170,7 +3299,7 @@ rebuild_relation_finish_concurrent(Relation NewHeap, Relation OldHeap,
* hold AccessExclusiveLock. (Quite some amount of WAL could have been
* written during the data copying and index creation.)
*/
- process_concurrent_changes(decoding_ctx, end_of_wal, &chgdst);
+ process_concurrent_changes(end_of_wal, &chgdst, false);
/*
* Acquire AccessExclusiveLock on the table, its TOAST relation (if there
@@ -3266,8 +3395,11 @@ rebuild_relation_finish_concurrent(Relation NewHeap, Relation OldHeap,
XLogFlush(wal_insert_ptr);
end_of_wal = GetFlushRecPtr(NULL);
- /* Apply the concurrent changes again. */
- process_concurrent_changes(decoding_ctx, end_of_wal, &chgdst);
+ /*
+ * Apply the concurrent changes again. Indicate that the decoding worker
+ * won't be needed anymore.
+ */
+ process_concurrent_changes(end_of_wal, &chgdst, true);
/* Remember info about rel before closing OldHeap */
relpersistence = OldHeap->rd_rel->relpersistence;
@@ -3380,3 +3512,488 @@ build_new_indexes(Relation NewHeap, Relation OldHeap, List *OldIndexes)
return result;
}
+
+/*
+ * Try to start a background worker to perform logical decoding of data
+ * changes applied to relation while REPACK CONCURRENTLY is copying its
+ * contents to a new table.
+ */
+static void
+start_decoding_worker(Oid relid)
+{
+ Size size;
+ dsm_segment *seg;
+ DecodingWorkerShared *shared;
+ shm_mq *mq;
+ shm_mq_handle *mqh;
+ BackgroundWorker bgw;
+
+ /* Setup shared memory. */
+ size = BUFFERALIGN(offsetof(DecodingWorkerShared, error_queue)) +
+ BUFFERALIGN(REPACK_ERROR_QUEUE_SIZE);
+ seg = dsm_create(size, 0);
+ shared = (DecodingWorkerShared *) dsm_segment_address(seg);
+ shared->lsn_upto = InvalidXLogRecPtr;
+ shared->done = false;
+ SharedFileSetInit(&shared->sfs, seg);
+ shared->sfs_valid = false;
+ shared->last_exported = -1;
+ SpinLockInit(&shared->mutex);
+ shared->dbid = MyDatabaseId;
+
+ /*
+ * This is the UserId set in cluster_rel(). Security context shouldn't be
+ * needed for decoding worker.
+ */
+ shared->roleid = GetUserId();
+ shared->relid = relid;
+ ConditionVariableInit(&shared->cv);
+ shared->backend_proc = MyProc;
+ shared->backend_pid = MyProcPid;
+ shared->backend_proc_number = MyProcNumber;
+
+ mq = shm_mq_create((char *) BUFFERALIGN(shared->error_queue),
+ REPACK_ERROR_QUEUE_SIZE);
+ shm_mq_set_receiver(mq, MyProc);
+ mqh = shm_mq_attach(mq, seg, NULL);
+
+ memset(&bgw, 0, sizeof(bgw));
+ snprintf(bgw.bgw_name, BGW_MAXLEN,
+ "REPACK decoding worker for relation \"%s\"",
+ get_rel_name(relid));
+ snprintf(bgw.bgw_type, BGW_MAXLEN, "REPACK decoding worker");
+ bgw.bgw_flags = BGWORKER_SHMEM_ACCESS |
+ BGWORKER_BACKEND_DATABASE_CONNECTION;
+ bgw.bgw_start_time = BgWorkerStart_RecoveryFinished;
+ bgw.bgw_restart_time = BGW_NEVER_RESTART;
+ snprintf(bgw.bgw_library_name, MAXPGPATH, "postgres");
+ snprintf(bgw.bgw_function_name, BGW_MAXLEN, "RepackWorkerMain");
+ bgw.bgw_main_arg = UInt32GetDatum(dsm_segment_handle(seg));
+ bgw.bgw_notify_pid = MyProcPid;
+
+ decoding_worker = palloc0_object(DecodingWorker);
+ if (!RegisterDynamicBackgroundWorker(&bgw, &decoding_worker->handle))
+ ereport(ERROR,
+ (errcode(ERRCODE_CONFIGURATION_LIMIT_EXCEEDED),
+ errmsg("out of background worker slots"),
+ errhint("You might need to increase \"%s\".", "max_worker_processes")));
+
+ decoding_worker->seg = seg;
+ decoding_worker->error_mqh = mqh;
+}
+
+/*
+ * Stop the decoding worker and cleanup the related resources.
+ *
+ * The worker stops on its own when it knows there is no more work to do, but
+ * we need to stop it explicitly at least on ERROR in the launching backend.
+ */
+static void
+stop_decoding_worker(void)
+{
+ BgwHandleStatus status;
+
+ /* Haven't reached the worker startup? */
+ if (decoding_worker == NULL)
+ return;
+
+ /* Could not register the worker? */
+ if (decoding_worker->handle == NULL)
+ return;
+
+ TerminateBackgroundWorker(decoding_worker->handle);
+ /* The worker should really exit before the REPACK command does. */
+ HOLD_INTERRUPTS();
+ status = WaitForBackgroundWorkerShutdown(decoding_worker->handle);
+ RESUME_INTERRUPTS();
+
+ if (status == BGWH_POSTMASTER_DIED)
+ ereport(FATAL,
+ (errcode(ERRCODE_ADMIN_SHUTDOWN),
+ errmsg("postmaster exited during REPACK command")));
+
+ shm_mq_detach(decoding_worker->error_mqh);
+
+ /*
+ * If we could not cancel the current sleep due to ERROR, do that before
+ * we detach from the shared memory the condition variable is located in.
+ * If we did not, the bgworker ERROR handling code would try and fail
+ * badly.
+ */
+ ConditionVariableCancelSleep();
+
+ dsm_detach(decoding_worker->seg);
+ pfree(decoding_worker);
+ decoding_worker = NULL;
+}
+
+/* Is this process a REPACK worker? */
+static bool is_repack_worker = false;
+
+static pid_t backend_pid;
+static ProcNumber backend_proc_number;
+
+/*
+ * See ParallelWorkerShutdown for details.
+ */
+static void
+RepackWorkerShutdown(int code, Datum arg)
+{
+ SendProcSignal(backend_pid,
+ PROCSIG_REPACK_MESSAGE,
+ backend_proc_number);
+
+ dsm_detach((dsm_segment *) DatumGetPointer(arg));
+}
+
+/* REPACK decoding worker entry point */
+void
+RepackWorkerMain(Datum main_arg)
+{
+ dsm_segment *seg;
+ DecodingWorkerShared *shared;
+ shm_mq *mq;
+ shm_mq_handle *mqh;
+
+ is_repack_worker = true;
+
+ /*
+ * Override the default bgworker_die() with die() so we can use
+ * CHECK_FOR_INTERRUPTS().
+ */
+ pqsignal(SIGTERM, die);
+ BackgroundWorkerUnblockSignals();
+
+ seg = dsm_attach(DatumGetUInt32(main_arg));
+ if (seg == NULL)
+ ereport(ERROR,
+ (errcode(ERRCODE_OBJECT_NOT_IN_PREREQUISITE_STATE),
+ errmsg("could not map dynamic shared memory segment")));
+
+ shared = (DecodingWorkerShared *) dsm_segment_address(seg);
+
+ /* Arrange to signal the leader if we exit. */
+ backend_pid = shared->backend_pid;
+ backend_proc_number = shared->backend_proc_number;
+ before_shmem_exit(RepackWorkerShutdown, PointerGetDatum(seg));
+
+ /*
+ * Join locking group - see the comments around the call of
+ * start_decoding_worker().
+ */
+ if (!BecomeLockGroupMember(shared->backend_proc, backend_pid))
+ /* The leader is not running anymore. */
+ return;
+
+ /*
+ * Setup a queue to send error messages to the backend that launched this
+ * worker.
+ */
+ mq = (shm_mq *) (char *) BUFFERALIGN(shared->error_queue);
+ shm_mq_set_sender(mq, MyProc);
+ mqh = shm_mq_attach(mq, seg, NULL);
+ pq_redirect_to_shm_mq(seg, mqh);
+ pq_set_parallel_leader(shared->backend_pid,
+ shared->backend_proc_number);
+
+ /* Connect to the database. */
+ BackgroundWorkerInitializeConnectionByOid(shared->dbid, shared->roleid, 0);
+
+ repack_worker_internal(seg);
+}
+
+static void
+repack_worker_internal(dsm_segment *seg)
+{
+ DecodingWorkerShared *shared;
+ LogicalDecodingContext *decoding_ctx;
+ SharedFileSet *sfs;
+ Snapshot snapshot;
+
+ /*
+ * Transaction is needed to open relation, and it also provides us with a
+ * resource owner.
+ */
+ StartTransactionCommand();
+
+ shared = (DecodingWorkerShared *) dsm_segment_address(seg);
+
+ /*
+ * Not sure the spinlock is needed here - the backend should not change
+ * anything in the shared memory until we have serialized the snapshot.
+ */
+ SpinLockAcquire(&shared->mutex);
+ Assert(XLogRecPtrIsInvalid(shared->lsn_upto));
+ Assert(!shared->sfs_valid);
+ sfs = &shared->sfs;
+ SpinLockRelease(&shared->mutex);
+
+ SharedFileSetAttach(sfs, seg);
+
+ /*
+ * Prepare to capture the concurrent data changes ourselves.
+ */
+ decoding_ctx = setup_logical_decoding(shared->relid);
+
+ /* Build the initial snapshot and export it. */
+ snapshot = SnapBuildInitialSnapshotForRepack(decoding_ctx->snapshot_builder);
+ export_initial_snapshot(snapshot, shared);
+
+ /*
+ * The worker already had to access some system catalogs during startup,
+ * and we even had to open the relation we are processing. Now that we're
+ * going to work with historic snapshots, the system caches must be
+ * invalidated.
+ */
+ InvalidateSystemCaches();
+
+ while (!decode_concurrent_changes(decoding_ctx, shared))
+ ;
+
+ /* Cleanup. */
+ cleanup_logical_decoding(decoding_ctx);
+ CommitTransactionCommand();
+}
+
+/*
+ * Make snapshot available to the backend that launched the decoding worker.
+ */
+static void
+export_initial_snapshot(Snapshot snapshot, DecodingWorkerShared *shared)
+{
+ char fname[MAXPGPATH];
+ BufFile *file;
+ Size snap_size;
+ char *snap_space;
+
+ snap_size = EstimateSnapshotSpace(snapshot);
+ snap_space = (char *) palloc(snap_size);
+ SerializeSnapshot(snapshot, snap_space);
+ FreeSnapshot(snapshot);
+
+ DecodingWorkerFileName(fname, shared->relid, shared->last_exported + 1);
+ file = BufFileCreateFileSet(&shared->sfs.fs, fname);
+ /* To make restoration easier, write the snapshot size first. */
+ BufFileWrite(file, &snap_size, sizeof(snap_size));
+ BufFileWrite(file, snap_space, snap_size);
+ pfree(snap_space);
+ BufFileClose(file);
+
+ /* Tell the backend that the file is available. */
+ SpinLockAcquire(&shared->mutex);
+ shared->sfs_valid = true;
+ shared->last_exported++;
+ SpinLockRelease(&shared->mutex);
+ ConditionVariableSignal(&shared->cv);
+}
+
+/*
+ * Get the initial snapshot from the decoding worker.
+ */
+static Snapshot
+get_initial_snapshot(DecodingWorker *worker)
+{
+ DecodingWorkerShared *shared;
+ char fname[MAXPGPATH];
+ BufFile *file;
+ Size snap_size;
+ char *snap_space;
+ Snapshot snapshot;
+
+ shared = (DecodingWorkerShared *) dsm_segment_address(worker->seg);
+
+ /*
+ * The worker needs to initialize the logical decoding, which usually
+ * takes some time. Therefore it makes sense to prepare for the sleep
+ * first.
+ */
+ ConditionVariablePrepareToSleep(&shared->cv);
+ for (;;)
+ {
+ bool valid;
+
+ SpinLockAcquire(&shared->mutex);
+ valid = shared->sfs_valid;
+ SpinLockRelease(&shared->mutex);
+
+ if (valid)
+ break;
+
+ ConditionVariableSleep(&shared->cv, WAIT_EVENT_REPACK_WORKER_EXPORT);
+ }
+ ConditionVariableCancelSleep();
+
+ /* Read the snapshot from a file. */
+ DecodingWorkerFileName(fname, shared->relid, shared->last_exported);
+ file = BufFileOpenFileSet(&shared->sfs.fs, fname, O_RDONLY, false);
+ BufFileReadExact(file, &snap_size, sizeof(snap_size));
+ snap_space = (char *) palloc(snap_size);
+ BufFileReadExact(file, snap_space, snap_size);
+ BufFileClose(file);
+
+ SpinLockAcquire(&shared->mutex);
+ shared->sfs_valid = false;
+ SpinLockRelease(&shared->mutex);
+
+ /* Restore it. */
+ snapshot = RestoreSnapshot(snap_space);
+ pfree(snap_space);
+
+ return snapshot;
+}
+
+bool
+IsRepackWorker(void)
+{
+ return is_repack_worker;
+}
+
+/*
+ * Handle receipt of an interrupt indicating a repack worker message.
+ *
+ * Note: this is called within a signal handler! All we can do is set
+ * a flag that will cause the next CHECK_FOR_INTERRUPTS() to invoke
+ * ProcessRepackMessages().
+ */
+void
+HandleRepackMessageInterrupt(void)
+{
+ InterruptPending = true;
+ RepackMessagePending = true;
+ SetLatch(MyLatch);
+}
+
+/*
+ * Process any queued protocol messages received from parallel workers.
+ */
+void
+ProcessRepackMessages(void)
+{
+ MemoryContext oldcontext;
+
+ static MemoryContext hpm_context = NULL;
+
+ /*
+ * Nothing to do if we haven't launched the worker yet or have already
+ * terminated it.
+ */
+ if (decoding_worker == NULL)
+ return;
+
+ /*
+ * This is invoked from ProcessInterrupts(), and since some of the
+ * functions it calls contain CHECK_FOR_INTERRUPTS(), there is a potential
+ * for recursive calls if more signals are received while this runs. It's
+ * unclear that recursive entry would be safe, and it doesn't seem useful
+ * even if it is safe, so let's block interrupts until done.
+ */
+ HOLD_INTERRUPTS();
+
+ /*
+ * Moreover, CurrentMemoryContext might be pointing almost anywhere. We
+ * don't want to risk leaking data into long-lived contexts, so let's do
+ * our work here in a private context that we can reset on each use.
+ */
+ if (hpm_context == NULL) /* first time through? */
+ hpm_context = AllocSetContextCreate(TopMemoryContext,
+ "ProcessParallelMessages",
+ ALLOCSET_DEFAULT_SIZES);
+ else
+ MemoryContextReset(hpm_context);
+
+ oldcontext = MemoryContextSwitchTo(hpm_context);
+
+ /* OK to process messages. Reset the flag saying there are more to do. */
+ RepackMessagePending = false;
+
+ /*
+ * Read as many messages as we can from each worker, but stop when no more
+ * messages can be read from the worker without blocking.
+ */
+ while (true)
+ {
+ shm_mq_result res;
+ Size nbytes;
+ void *data;
+
+ res = shm_mq_receive(decoding_worker->error_mqh, &nbytes,
+ &data, true);
+ if (res == SHM_MQ_WOULD_BLOCK)
+ break;
+ else if (res == SHM_MQ_SUCCESS)
+ {
+ StringInfoData msg;
+
+ initStringInfo(&msg);
+ appendBinaryStringInfo(&msg, data, nbytes);
+ ProcessRepackMessage(&msg);
+ pfree(msg.data);
+ }
+ else
+ {
+ /*
+ * The decoding worker is special in that it exits as soon as it
+ * has its work done. Thus the DETACHED result code is fine.
+ */
+ Assert(res = SHM_MQ_DETACHED);
+
+ break;
+ }
+ }
+
+ MemoryContextSwitchTo(oldcontext);
+
+ /* Might as well clear the context on our way out */
+ MemoryContextReset(hpm_context);
+
+ RESUME_INTERRUPTS();
+}
+
+/*
+ * Process a single protocol message received from a single parallel worker.
+ */
+static void
+ProcessRepackMessage(StringInfo msg)
+{
+ char msgtype;
+
+ msgtype = pq_getmsgbyte(msg);
+
+ switch (msgtype)
+ {
+ case PqMsg_ErrorResponse:
+ case PqMsg_NoticeResponse:
+ {
+ ErrorData edata;
+
+ /* Parse ErrorResponse or NoticeResponse. */
+ pq_parse_errornotice(msg, &edata);
+
+ /* Death of a worker isn't enough justification for suicide. */
+ edata.elevel = Min(edata.elevel, ERROR);
+
+ /*
+ * If desired, add a context line to show that this is a
+ * message propagated from a parallel worker. Otherwise, it
+ * can sometimes be confusing to understand what actually
+ * happened.
+ */
+ if (edata.context)
+ edata.context = psprintf("%s\n%s", edata.context,
+ _("decoding worker"));
+ else
+ edata.context = pstrdup(_("decoding worker"));
+
+ /* Rethrow error or print notice. */
+ ThrowErrorData(&edata);
+
+ break;
+ }
+
+ default:
+ {
+ elog(ERROR, "unrecognized message type received from decoding worker: %c (message length %d bytes)",
+ msgtype, msg->len);
+ }
+ }
+}
diff --git a/src/backend/libpq/pqmq.c b/src/backend/libpq/pqmq.c
index 2b75de0ddef..28a5a400fb1 100644
--- a/src/backend/libpq/pqmq.c
+++ b/src/backend/libpq/pqmq.c
@@ -14,6 +14,7 @@
#include "postgres.h"
#include "access/parallel.h"
+#include "commands/cluster.h"
#include "libpq/libpq.h"
#include "libpq/pqformat.h"
#include "libpq/pqmq.h"
@@ -175,6 +176,10 @@ mq_putmessage(char msgtype, const char *s, size_t len)
SendProcSignal(pq_mq_parallel_leader_pid,
PROCSIG_PARALLEL_APPLY_MESSAGE,
pq_mq_parallel_leader_proc_number);
+ else if (IsRepackWorker())
+ SendProcSignal(pq_mq_parallel_leader_pid,
+ PROCSIG_REPACK_MESSAGE,
+ pq_mq_parallel_leader_proc_number);
else
{
Assert(IsParallelWorker());
diff --git a/src/backend/postmaster/bgworker.c b/src/backend/postmaster/bgworker.c
index 142a02eb5e9..b368990e90b 100644
--- a/src/backend/postmaster/bgworker.c
+++ b/src/backend/postmaster/bgworker.c
@@ -13,6 +13,7 @@
#include "postgres.h"
#include "access/parallel.h"
+#include "commands/cluster.h"
#include "libpq/pqsignal.h"
#include "miscadmin.h"
#include "pgstat.h"
@@ -135,6 +136,9 @@ static const struct
},
{
"SequenceSyncWorkerMain", SequenceSyncWorkerMain
+ },
+ {
+ "RepackWorkerMain", RepackWorkerMain
}
};
diff --git a/src/backend/replication/logical/logical.c b/src/backend/replication/logical/logical.c
index 866f92cf799..85323ba61e2 100644
--- a/src/backend/replication/logical/logical.c
+++ b/src/backend/replication/logical/logical.c
@@ -205,7 +205,11 @@ StartupDecodingContext(List *output_plugin_options,
ctx->slot = slot;
- ctx->reader = XLogReaderAllocate(wal_segment_size, NULL, xl_routine, ctx);
+ /*
+ * TODO A separate patch for PG core, unless there's really a reason to
+ * pass ctx for private_data (May extensions expect ctx?).
+ */
+ ctx->reader = XLogReaderAllocate(wal_segment_size, NULL, xl_routine, NULL);
if (!ctx->reader)
ereport(ERROR,
(errcode(ERRCODE_OUT_OF_MEMORY),
diff --git a/src/backend/replication/pgoutput_repack/pgoutput_repack.c b/src/backend/replication/pgoutput_repack/pgoutput_repack.c
index c8930640a0d..fb9956d392d 100644
--- a/src/backend/replication/pgoutput_repack/pgoutput_repack.c
+++ b/src/backend/replication/pgoutput_repack/pgoutput_repack.c
@@ -168,17 +168,13 @@ store_change(LogicalDecodingContext *ctx, ConcurrentChangeKind kind,
HeapTuple tuple)
{
RepackDecodingState *dstate;
- char *change_raw;
- ConcurrentChange change;
+ char kind_byte = (char) kind;
bool flattened = false;
- Size size;
- Datum values[1];
- bool isnull[1];
- char *dst;
dstate = (RepackDecodingState *) ctx->output_writer_private;
- size = VARHDRSZ + SizeOfConcurrentChange;
+ /* Store the change kind. */
+ BufFileWrite(dstate->file, &kind_byte, 1);
/*
* ReorderBufferCommit() stores the TOAST chunks in its private memory
@@ -195,46 +191,12 @@ store_change(LogicalDecodingContext *ctx, ConcurrentChangeKind kind,
tuple = toast_flatten_tuple(tuple, dstate->tupdesc);
flattened = true;
}
+ /* Store the tuple size ... */
+ BufFileWrite(dstate->file, &tuple->t_len, sizeof(tuple->t_len));
+ /* ... and the tuple itself. */
+ BufFileWrite(dstate->file, tuple->t_data, tuple->t_len);
- size += tuple->t_len;
- if (size >= MaxAllocSize)
- elog(ERROR, "Change is too big.");
-
- /* Construct the change. */
- change_raw = (char *) palloc0(size);
- SET_VARSIZE(change_raw, size);
-
- /*
- * Since the varlena alignment might not be sufficient for the structure,
- * set the fields in a local instance and remember where it should
- * eventually be copied.
- */
- change.kind = kind;
- dst = (char *) VARDATA(change_raw);
-
- /*
- * Copy the tuple.
- *
- * Note: change->tup_data.t_data must be fixed on retrieval!
- */
- memcpy(&change.tup_data, tuple, sizeof(HeapTupleData));
- memcpy(dst, &change, SizeOfConcurrentChange);
- dst += SizeOfConcurrentChange;
- memcpy(dst, tuple->t_data, tuple->t_len);
-
- /* The data has been copied. */
+ /* Free the flat copy if created above. */
if (flattened)
pfree(tuple);
-
- /* Store as tuple of 1 bytea column. */
- values[0] = PointerGetDatum(change_raw);
- isnull[0] = false;
- tuplestore_putvalues(dstate->tstore, dstate->tupdesc_change,
- values, isnull);
-
- /* Accounting. */
- dstate->nchanges++;
-
- /* Cleanup. */
- pfree(change_raw);
}
diff --git a/src/backend/storage/ipc/procsignal.c b/src/backend/storage/ipc/procsignal.c
index 087821311cc..af12144795b 100644
--- a/src/backend/storage/ipc/procsignal.c
+++ b/src/backend/storage/ipc/procsignal.c
@@ -19,6 +19,7 @@
#include "access/parallel.h"
#include "commands/async.h"
+#include "commands/cluster.h"
#include "miscadmin.h"
#include "pgstat.h"
#include "port/pg_bitutils.h"
@@ -694,6 +695,9 @@ procsignal_sigusr1_handler(SIGNAL_ARGS)
if (CheckProcSignal(PROCSIG_PARALLEL_APPLY_MESSAGE))
HandleParallelApplyMessageInterrupt();
+ if (CheckProcSignal(PROCSIG_REPACK_MESSAGE))
+ HandleRepackMessageInterrupt();
+
if (CheckProcSignal(PROCSIG_RECOVERY_CONFLICT_DATABASE))
HandleRecoveryConflictInterrupt(PROCSIG_RECOVERY_CONFLICT_DATABASE);
diff --git a/src/backend/tcop/postgres.c b/src/backend/tcop/postgres.c
index 7dd75a490aa..4a4858882f0 100644
--- a/src/backend/tcop/postgres.c
+++ b/src/backend/tcop/postgres.c
@@ -36,6 +36,7 @@
#include "access/xact.h"
#include "catalog/pg_type.h"
#include "commands/async.h"
+#include "commands/cluster.h"
#include "commands/event_trigger.h"
#include "commands/explain_state.h"
#include "commands/prepare.h"
@@ -3541,6 +3542,9 @@ ProcessInterrupts(void)
if (ParallelApplyMessagePending)
ProcessParallelApplyMessages();
+
+ if (RepackMessagePending)
+ ProcessRepackMessages();
}
/*
diff --git a/src/backend/utils/activity/wait_event_names.txt b/src/backend/utils/activity/wait_event_names.txt
index f39830dbb34..cbcc8550960 100644
--- a/src/backend/utils/activity/wait_event_names.txt
+++ b/src/backend/utils/activity/wait_event_names.txt
@@ -62,6 +62,7 @@ LOGICAL_APPLY_MAIN "Waiting in main loop of logical replication apply process."
LOGICAL_LAUNCHER_MAIN "Waiting in main loop of logical replication launcher process."
LOGICAL_PARALLEL_APPLY_MAIN "Waiting in main loop of logical replication parallel apply process."
RECOVERY_WAL_STREAM "Waiting in main loop of startup process for WAL to arrive, during streaming recovery."
+REPACK_WORKER_MAIN "Waiting in main loop of REPACK decoding worker process."
REPLICATION_SLOTSYNC_MAIN "Waiting in main loop of slot sync worker."
REPLICATION_SLOTSYNC_SHUTDOWN "Waiting for slot sync worker to shut down."
SYSLOGGER_MAIN "Waiting in main loop of syslogger process."
@@ -153,6 +154,7 @@ RECOVERY_CONFLICT_SNAPSHOT "Waiting for recovery conflict resolution for a vacuu
RECOVERY_CONFLICT_TABLESPACE "Waiting for recovery conflict resolution for dropping a tablespace."
RECOVERY_END_COMMAND "Waiting for <xref linkend="guc-recovery-end-command"/> to complete."
RECOVERY_PAUSE "Waiting for recovery to be resumed."
+REPACK_WORKER_EXPORT "Waiting for decoding worker to export a new output file."
REPLICATION_ORIGIN_DROP "Waiting for a replication origin to become inactive so it can be dropped."
REPLICATION_SLOT_DROP "Waiting for a replication slot to become inactive so it can be dropped."
RESTORE_COMMAND "Waiting for <xref linkend="guc-restore-command"/> to complete."
diff --git a/src/include/access/tableam.h b/src/include/access/tableam.h
index d8f76d325f9..2b983abce3e 100644
--- a/src/include/access/tableam.h
+++ b/src/include/access/tableam.h
@@ -22,7 +22,6 @@
#include "access/xact.h"
#include "commands/vacuum.h"
#include "executor/tuptable.h"
-#include "replication/logical.h"
#include "storage/read_stream.h"
#include "utils/rel.h"
#include "utils/snapshot.h"
@@ -631,7 +630,6 @@ typedef struct TableAmRoutine
bool use_sort,
TransactionId OldestXmin,
Snapshot snapshot,
- LogicalDecodingContext *decoding_ctx,
TransactionId *xid_cutoff,
MultiXactId *multi_cutoff,
double *num_tuples,
@@ -1652,7 +1650,7 @@ table_relation_copy_data(Relation rel, const RelFileLocator *newrlocator)
* - snapshot - if != NULL, ignore data changes done by transactions that this
* (MVCC) snapshot considers still in-progress or in the future.
* - decoding_ctx - logical decoding context, to capture concurrent data
- * changes.
+ * changes. NULL if background worker takes care of the decoding.
*
* Output parameters:
* - *xid_cutoff - rel's new relfrozenxid value, may be invalid
@@ -1666,7 +1664,6 @@ table_relation_copy_for_cluster(Relation OldTable, Relation NewTable,
bool use_sort,
TransactionId OldestXmin,
Snapshot snapshot,
- LogicalDecodingContext *decoding_ctx,
TransactionId *xid_cutoff,
MultiXactId *multi_cutoff,
double *num_tuples,
@@ -1675,7 +1672,7 @@ table_relation_copy_for_cluster(Relation OldTable, Relation NewTable,
{
OldTable->rd_tableam->relation_copy_for_cluster(OldTable, NewTable, OldIndex,
use_sort, OldestXmin,
- snapshot, decoding_ctx,
+ snapshot,
xid_cutoff, multi_cutoff,
num_tuples, tups_vacuumed,
tups_recently_dead);
diff --git a/src/include/commands/cluster.h b/src/include/commands/cluster.h
index b43a1740053..0ac70ec30d7 100644
--- a/src/include/commands/cluster.h
+++ b/src/include/commands/cluster.h
@@ -17,6 +17,7 @@
#include "nodes/parsenodes.h"
#include "parser/parse_node.h"
#include "replication/logical.h"
+#include "storage/buffile.h"
#include "storage/lock.h"
#include "storage/relfilelocator.h"
#include "utils/relcache.h"
@@ -47,6 +48,9 @@ typedef struct ClusterParams
extern RelFileLocator repacked_rel_locator;
extern RelFileLocator repacked_rel_toast_locator;
+/*
+ * Stored as a single byte in the output file.
+ */
typedef enum
{
CHANGE_INSERT,
@@ -55,68 +59,30 @@ typedef enum
CHANGE_DELETE
} ConcurrentChangeKind;
-typedef struct ConcurrentChange
-{
- /* See the enum above. */
- ConcurrentChangeKind kind;
-
- /*
- * The actual tuple.
- *
- * The tuple data follows the ConcurrentChange structure. Before use make
- * sure the tuple is correctly aligned (ConcurrentChange can be stored as
- * bytea) and that tuple->t_data is fixed.
- */
- HeapTupleData tup_data;
-} ConcurrentChange;
-
-#define SizeOfConcurrentChange (offsetof(ConcurrentChange, tup_data) + \
- sizeof(HeapTupleData))
-
/*
* Logical decoding state.
*
- * Here we store the data changes that we decode from WAL while the table
- * contents is being copied to a new storage. Also the necessary metadata
- * needed to apply these changes to the table is stored here.
+ * The output plugin uses it to store the data changes that it decodes from
+ * WAL while the table contents is being copied to a new storage.
*/
typedef struct RepackDecodingState
{
/* The relation whose changes we're decoding. */
Oid relid;
- /* Replication slot name. */
- NameData slotname;
-
- /*
- * Decoded changes are stored here. Although we try to avoid excessive
- * batches, it can happen that the changes need to be stored to disk. The
- * tuplestore does this transparently.
- */
- Tuplestorestate *tstore;
-
- /* The current number of changes in tstore. */
- double nchanges;
-
- /*
- * Descriptor to store the ConcurrentChange structure serialized (bytea).
- * We can't store the tuple directly because tuplestore only supports
- * minimum tuple and we may need to transfer OID system column from the
- * output plugin. Also we need to transfer the change kind, so it's better
- * to put everything in the structure than to use 2 tuplestores "in
- * parallel".
- */
- TupleDesc tupdesc_change;
-
- /* Tuple descriptor needed to update indexes. */
+ /* Tuple descriptor of the relation being processed. */
TupleDesc tupdesc;
- /* Slot to retrieve data from tstore. */
- TupleTableSlot *tsslot;
-
- ResourceOwner resowner;
+ /* The current output file. */
+ BufFile *file;
} RepackDecodingState;
+extern PGDLLIMPORT volatile sig_atomic_t RepackMessagePending;
+
+extern bool IsRepackWorker(void);
+extern void HandleRepackMessageInterrupt(void);
+extern void ProcessRepackMessages(void);
+
extern void ExecRepack(ParseState *pstate, RepackStmt *stmt, bool isTopLevel);
extern void cluster_rel(RepackCommand command, Relation OldHeap, Oid indexOid,
@@ -125,9 +91,6 @@ extern void check_index_is_clusterable(Relation OldHeap, Oid indexOid,
LOCKMODE lockmode);
extern void mark_index_clustered(Relation rel, Oid indexOid, bool is_internal);
-extern void repack_decode_concurrent_changes(LogicalDecodingContext *ctx,
- XLogRecPtr end_of_wal);
-
extern Oid make_new_heap(Oid OIDOldHeap, Oid NewTableSpace, Oid NewAccessMethod,
char relpersistence, LOCKMODE lockmode);
extern void finish_heap_swap(Oid OIDOldHeap, Oid OIDNewHeap,
@@ -140,4 +103,5 @@ extern void finish_heap_swap(Oid OIDOldHeap, Oid OIDNewHeap,
MultiXactId cutoffMulti,
char newrelpersistence);
+extern void RepackWorkerMain(Datum main_arg);
#endif /* CLUSTER_H */
diff --git a/src/include/storage/procsignal.h b/src/include/storage/procsignal.h
index afeeb1ca019..c0a66516b66 100644
--- a/src/include/storage/procsignal.h
+++ b/src/include/storage/procsignal.h
@@ -36,6 +36,7 @@ typedef enum
PROCSIG_BARRIER, /* global barrier interrupt */
PROCSIG_LOG_MEMORY_CONTEXT, /* ask backend to log the memory contexts */
PROCSIG_PARALLEL_APPLY_MESSAGE, /* Message from parallel apply workers */
+ PROCSIG_REPACK_MESSAGE, /* Message from repack worker */
/* Recovery conflict reasons */
PROCSIG_RECOVERY_CONFLICT_FIRST,
diff --git a/src/tools/pgindent/typedefs.list b/src/tools/pgindent/typedefs.list
index 3139b14e85f..35344910f65 100644
--- a/src/tools/pgindent/typedefs.list
+++ b/src/tools/pgindent/typedefs.list
@@ -488,7 +488,6 @@ CompressFileHandle
CompressionLocation
CompressorState
ComputeXidHorizonsResult
-ConcurrentChange
ConcurrentChangeKind
ConditionVariable
ConditionVariableMinimallyPadded
@@ -629,6 +628,9 @@ DeclareCursorStmt
DecodedBkpBlock
DecodedXLogRecord
DecodingOutputState
+DecodingWorker
+DecodingWorkerShared
+DecodingWorkerState
DefElem
DefElemAction
DefaultACLInfo
--
2.47.3
--=-=-=--
^ permalink raw reply [nested|flat] 213+ messages in thread
end of thread, other threads:[~2025-12-09 18:44 UTC | newest]
Thread overview: 213+ messages (download: mbox mbox.gz follow: Atom feed)
-- links below jump to the message on this page --
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2024-10-16 18:21 [PATCH v4 2/8] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-06-17 09:47 [PATCH v5 05/10] Address space reservation for shared memory Dmitrii Dolgov <9erthalion6@gmail.com>
2025-12-09 18:44 [PATCH 5/5] Use background worker to do logical decoding. Antonin Houska <ah@cybertec.at>
This inbox is served by agora; see mirroring instructions
for how to clone and mirror all data and code used for this inbox