agora inbox for [email protected]
help / color / mirror / Atom feedTiming of relcache inval at parallel worker init
268+ messages / 4 participants
[nested] [flat]
* Timing of relcache inval at parallel worker init
@ 2020-10-17 11:53 Noah Misch <[email protected]>
2020-10-20 08:35 ` Re: Timing of relcache inval at parallel worker init Kyotaro Horiguchi <[email protected]>
2020-10-21 15:31 ` Re: Timing of relcache inval at parallel worker init Robert Haas <[email protected]>
0 siblings, 2 replies; 268+ messages in thread
From: Noah Misch @ 2020-10-17 11:53 UTC (permalink / raw)
To: pgsql-hackers
While reviewing what became commit fe4d022, I was surprised at the sequence of
relfilenode values that RelationInitPhysicalAddr() computed for pg_class,
during ParallelWorkerMain(), when running the last command of this recipe:
begin;
cluster pg_class using pg_class_oid_index;
set force_parallel_mode = 'regress';
values (1);
There's $OLD_NODE (relfilenode in the committed relation map) and $NEW_NODE
(relfilenode in this transaction's active_local_updates). The worker performs
RelationInitPhysicalAddr(pg_class) four times:
1) $OLD_NODE in BackgroundWorkerInitializeConnectionByOid().
2) $OLD_NODE in RelationCacheInvalidate() directly.
3) $OLD_NODE in RelationReloadNailed(), indirectly via RelationCacheInvalidate().
4) $NEW_NODE indirectly as part of the executor running the query.
I did expect $OLD_NODE in (1), since ParallelWorkerMain() calls
BackgroundWorkerInitializeConnectionByOid() before
StartParallelWorkerTransaction(). I expected $NEW_NODE in (2) and (3); that
didn't happen, because ParallelWorkerMain() calls InvalidateSystemCaches()
before RestoreRelationMap(). Let's move InvalidateSystemCaches() later.
Invalidation should follow any worker initialization step that changes the
results of relcache validation; otherwise, we'd need to ensure the
InvalidateSystemCaches() will not validate any relcache entry. Invalidation
should precede any step that reads from a cache; otherwise, we'd need to redo
that step after inval. (Currently, no step reads from a cache.) Many steps,
e.g. AttachSerializableXact(), have no effect on relcache validation, so it's
arbitrary whether they happen before or after inval. I'm putting inval as
late as possible, because I think it's easier to confirm that a step doesn't
read from a cache than to confirm that a step doesn't affect relcache
validation. An also-reasonable alternative would be to move inval and its
prerequisites as early as possible.
For reasons described in the attached commit message, this doesn't have
user-visible consequences today. Innocent-looking relcache.c changes might
upheave that, so I'm proposing this on robustness grounds. No need to
back-patch.
Author: Noah Misch <[email protected]>
Commit: Noah Misch <[email protected]>
In parallel workers, invalidate after restoring relmapper.
A relcache entry is no fresher than the relmapper data used to build it.
The "rd_refcnt <= 1" in RelationReloadNailed() prevented user-visible
consequences by leaving every relation !rd_isvalid at the end of this
InvalidateSystemCaches(). Any subsequent relation usage first validated
the relation, drawing on correct relmapper data. Even so, reduce
fragility by delaying invalidation until we initialize all state
essential to rebuilding a relcache entry.
Reviewed by FIXME.
Discussion: https://postgr.es/m/FIXME
diff --git a/src/backend/access/transam/parallel.c b/src/backend/access/transam/parallel.c
index b042696..566ae87 100644
--- a/src/backend/access/transam/parallel.c
+++ b/src/backend/access/transam/parallel.c
@@ -1417,12 +1417,6 @@ ParallelWorkerMain(Datum main_arg)
PushActiveSnapshot(RestoreSnapshot(asnapspace));
/*
- * We've changed which tuples we can see, and must therefore invalidate
- * system caches.
- */
- InvalidateSystemCaches();
-
- /*
* Restore current role id. Skip verifying whether session user is
* allowed to become this role and blindly restore the leader's state for
* current role.
@@ -1458,6 +1452,12 @@ ParallelWorkerMain(Datum main_arg)
AttachSerializableXact(fps->serializable_xact_handle);
/*
+ * We've changed tuple visibility and relmapper state, so discard all
+ * provisional state derived from system catalogs.
+ */
+ InvalidateSystemCaches();
+
+ /*
* We've initialized all of our state now; nothing should change
* hereafter.
*/
Attachments:
[text/plain] parallel-worker-inval-timing-v1.patch (1.7K, ../../[email protected]/2-parallel-worker-inval-timing-v1.patch)
download | inline diff:
Author: Noah Misch <[email protected]>
Commit: Noah Misch <[email protected]>
In parallel workers, invalidate after restoring relmapper.
A relcache entry is no fresher than the relmapper data used to build it.
The "rd_refcnt <= 1" in RelationReloadNailed() prevented user-visible
consequences by leaving every relation !rd_isvalid at the end of this
InvalidateSystemCaches(). Any subsequent relation usage first validated
the relation, drawing on correct relmapper data. Even so, reduce
fragility by delaying invalidation until we initialize all state
essential to rebuilding a relcache entry.
Reviewed by FIXME.
Discussion: https://postgr.es/m/FIXME
diff --git a/src/backend/access/transam/parallel.c b/src/backend/access/transam/parallel.c
index b042696..566ae87 100644
--- a/src/backend/access/transam/parallel.c
+++ b/src/backend/access/transam/parallel.c
@@ -1417,12 +1417,6 @@ ParallelWorkerMain(Datum main_arg)
PushActiveSnapshot(RestoreSnapshot(asnapspace));
/*
- * We've changed which tuples we can see, and must therefore invalidate
- * system caches.
- */
- InvalidateSystemCaches();
-
- /*
* Restore current role id. Skip verifying whether session user is
* allowed to become this role and blindly restore the leader's state for
* current role.
@@ -1458,6 +1452,12 @@ ParallelWorkerMain(Datum main_arg)
AttachSerializableXact(fps->serializable_xact_handle);
/*
+ * We've changed tuple visibility and relmapper state, so discard all
+ * provisional state derived from system catalogs.
+ */
+ InvalidateSystemCaches();
+
+ /*
* We've initialized all of our state now; nothing should change
* hereafter.
*/
^ permalink raw reply [nested|flat] 268+ messages in thread
* Re: Timing of relcache inval at parallel worker init
2020-10-17 11:53 Timing of relcache inval at parallel worker init Noah Misch <[email protected]>
@ 2020-10-20 08:35 ` Kyotaro Horiguchi <[email protected]>
1 sibling, 0 replies; 268+ messages in thread
From: Kyotaro Horiguchi @ 2020-10-20 08:35 UTC (permalink / raw)
To: [email protected]; +Cc: pgsql-hackers
At Sat, 17 Oct 2020 04:53:06 -0700, Noah Misch <[email protected]> wrote in
> While reviewing what became commit fe4d022, I was surprised at the sequence of
> relfilenode values that RelationInitPhysicalAddr() computed for pg_class,
> during ParallelWorkerMain(), when running the last command of this recipe:
>
> begin;
> cluster pg_class using pg_class_oid_index;
> set force_parallel_mode = 'regress';
> values (1);
>
> There's $OLD_NODE (relfilenode in the committed relation map) and $NEW_NODE
> (relfilenode in this transaction's active_local_updates). The worker performs
> RelationInitPhysicalAddr(pg_class) four times:
>
> 1) $OLD_NODE in BackgroundWorkerInitializeConnectionByOid().
> 2) $OLD_NODE in RelationCacheInvalidate() directly.
> 3) $OLD_NODE in RelationReloadNailed(), indirectly via RelationCacheInvalidate().
> 4) $NEW_NODE indirectly as part of the executor running the query.
>
> I did expect $OLD_NODE in (1), since ParallelWorkerMain() calls
> BackgroundWorkerInitializeConnectionByOid() before
> StartParallelWorkerTransaction(). I expected $NEW_NODE in (2) and (3); that
> didn't happen, because ParallelWorkerMain() calls InvalidateSystemCaches()
> before RestoreRelationMap(). Let's move InvalidateSystemCaches() later.
> Invalidation should follow any worker initialization step that changes the
> results of relcache validation; otherwise, we'd need to ensure the
> InvalidateSystemCaches() will not validate any relcache entry. Invalidation
> should precede any step that reads from a cache; otherwise, we'd need to redo
> that step after inval. (Currently, no step reads from a cache.) Many steps,
> e.g. AttachSerializableXact(), have no effect on relcache validation, so it's
> arbitrary whether they happen before or after inval. I'm putting inval as
I agree to both the discussions.
> late as possible, because I think it's easier to confirm that a step doesn't
> read from a cache than to confirm that a step doesn't affect relcache
> validation. An also-reasonable alternative would be to move inval and its
> prerequisites as early as possible.
The steps became moved before the invalidation by this patch seems to
be in the lower layer than snapshot, so it seems to be reasonable.
> For reasons described in the attached commit message, this doesn't have
> user-visible consequences today. Innocent-looking relcache.c changes might
> upheave that, so I'm proposing this on robustness grounds. No need to
> back-patch.
I'm not sure about the necessity but lower-to-upper initialization
order is neat. I agree about not back-patching.
regards.
--
Kyotaro Horiguchi
NTT Open Source Software Center
^ permalink raw reply [nested|flat] 268+ messages in thread
* Re: Timing of relcache inval at parallel worker init
2020-10-17 11:53 Timing of relcache inval at parallel worker init Noah Misch <[email protected]>
@ 2020-10-21 15:31 ` Robert Haas <[email protected]>
2020-10-24 15:29 ` Re: Timing of relcache inval at parallel worker init Noah Misch <[email protected]>
1 sibling, 1 reply; 268+ messages in thread
From: Robert Haas @ 2020-10-21 15:31 UTC (permalink / raw)
To: Noah Misch <[email protected]>; +Cc: pgsql-hackers
On Sat, Oct 17, 2020 at 7:53 AM Noah Misch <[email protected]> wrote:
> Let's move InvalidateSystemCaches() later.
> Invalidation should follow any worker initialization step that changes the
> results of relcache validation; otherwise, we'd need to ensure the
> InvalidateSystemCaches() will not validate any relcache entry.
The thinking behind the current placement was this: just before the
call to InvalidateSystemCaches(), we restore the active and
transaction snapshots. I think that we must now flush the caches
before anyone does any more lookups; otherwise, they might get wrong
answers. So, putting this code later makes the assumption that no such
lookups happen meanwhile. That feels a little risky to me; at the very
least, it should be clearly spelled out in the comments that no system
cache lookups can happen in the functions we call in the interim.
Would it be obvious to a future developer that e.g.
RestoreEnumBlacklist cannot perform any syscache lookups? It doesn't
seem so to me.
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company
^ permalink raw reply [nested|flat] 268+ messages in thread
* Re: Timing of relcache inval at parallel worker init
2020-10-17 11:53 Timing of relcache inval at parallel worker init Noah Misch <[email protected]>
2020-10-21 15:31 ` Re: Timing of relcache inval at parallel worker init Robert Haas <[email protected]>
@ 2020-10-24 15:29 ` Noah Misch <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Noah Misch @ 2020-10-24 15:29 UTC (permalink / raw)
To: Robert Haas <[email protected]>; +Cc: pgsql-hackers
On Wed, Oct 21, 2020 at 11:31:54AM -0400, Robert Haas wrote:
> On Sat, Oct 17, 2020 at 7:53 AM Noah Misch <[email protected]> wrote:
> > Let's move InvalidateSystemCaches() later.
> > Invalidation should follow any worker initialization step that changes the
> > results of relcache validation; otherwise, we'd need to ensure the
> > InvalidateSystemCaches() will not validate any relcache entry.
>
> The thinking behind the current placement was this: just before the
> call to InvalidateSystemCaches(), we restore the active and
> transaction snapshots. I think that we must now flush the caches
> before anyone does any more lookups; otherwise, they might get wrong
> answers. So, putting this code later makes the assumption that no such
> lookups happen meanwhile. That feels a little risky to me; at the very
> least, it should be clearly spelled out in the comments that no system
> cache lookups can happen in the functions we call in the interim.
My comment edits did attempt that. I could enlarge those comments, at the
risk of putting undue weight on the topic. One could also arrange for an
assertion failure if something takes a snapshot in the unwelcome period,
between StartParallelWorkerTransaction() and InvalidateSystemCaches().
Looking closer, we have live bugs from lookups during RestoreGUCState():
$ echo "begin; create user alice; set session authorization alice; set force_parallel_mode = regress; select 1" | psql -X
BEGIN
CREATE ROLE
SET
SET
ERROR: role "alice" does not exist
CONTEXT: while setting parameter "session_authorization" to "alice"
$ echo "begin; create text search configuration foo (copy=english); set default_text_search_config = foo; set force_parallel_mode = regress; select 1" | psql -X
BEGIN
CREATE TEXT SEARCH CONFIGURATION
SET
SET
ERROR: invalid value for parameter "default_text_search_config": "public.foo"
CONTEXT: while setting parameter "default_text_search_config" to "public.foo"
I gather $SUBJECT is the tip of an iceberg, so I'm withdrawing the patch and
abandoning the topic.
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
* [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source
@ 2025-07-27 15:48 Lukas Fittl <[email protected]>
0 siblings, 0 replies; 268+ messages in thread
From: Lukas Fittl @ 2025-07-27 15:48 UTC (permalink / raw)
In passing also reduce the per-loop overhead caused by repeated divisions
in INSTR_TIME_GET_NANOSEC when the ticks variable has become very large,
instead diff first and then turn it into nanosecs.
---
src/bin/pg_test_timing/pg_test_timing.c | 59 +++++++++++++++++++------
src/include/portability/instr_time.h | 30 ++++++++-----
2 files changed, 65 insertions(+), 24 deletions(-)
diff --git a/src/bin/pg_test_timing/pg_test_timing.c b/src/bin/pg_test_timing/pg_test_timing.c
index a5621251afc..b77ef2063b6 100644
--- a/src/bin/pg_test_timing/pg_test_timing.c
+++ b/src/bin/pg_test_timing/pg_test_timing.c
@@ -16,6 +16,7 @@ static const char *progname;
static unsigned int test_duration = 3;
static double max_rprct = 99.99;
+static bool fast_timing = false;
/* record duration in powers of 2 nanoseconds */
static long long int histogram[32];
@@ -56,6 +57,7 @@ handle_args(int argc, char *argv[])
static struct option long_options[] = {
{"duration", required_argument, NULL, 'd'},
{"cutoff", required_argument, NULL, 'c'},
+ {"fast", no_argument, NULL, 'f'},
{NULL, 0, NULL, 0}
};
@@ -68,7 +70,7 @@ handle_args(int argc, char *argv[])
{
if (strcmp(argv[1], "--help") == 0 || strcmp(argv[1], "-?") == 0)
{
- printf(_("Usage: %s [-d DURATION] [-c CUTOFF]\n"), progname);
+ printf(_("Usage: %s [-d DURATION] [-c CUTOFF] [--fast]\n"), progname);
exit(0);
}
if (strcmp(argv[1], "--version") == 0 || strcmp(argv[1], "-V") == 0)
@@ -78,7 +80,7 @@ handle_args(int argc, char *argv[])
}
}
- while ((option = getopt_long(argc, argv, "d:c:",
+ while ((option = getopt_long(argc, argv, "d:c:f:",
long_options, &optindex)) != -1)
{
switch (option)
@@ -125,6 +127,10 @@ handle_args(int argc, char *argv[])
}
break;
+ case 'f':
+ fast_timing = true;
+ break;
+
default:
fprintf(stderr, _("Try \"%s --help\" for more information.\n"),
progname);
@@ -155,11 +161,31 @@ test_timing(unsigned int duration)
uint64 total_time;
int64 time_elapsed = 0;
uint64 loop_count = 0;
- uint64 prev,
- cur;
instr_time start_time,
end_time,
- temp;
+ prev,
+ cur;
+ char *time_source = NULL;
+ bool fast_timing_used = false;
+
+ INSTR_TIME_INITIALIZE();
+
+#if !defined(WIN32) && defined(__x86_64__) && defined(__linux__)
+ if (fast_timing && has_rdtsc)
+ {
+ time_source = "RDTSC";
+ fast_timing_used = true;
+ }
+ else if (has_rdtscp)
+ time_source = "RDTSCP";
+ else
+ time_source = PG_INSTR_CLOCK_NAME;
+#else
+ time_source = PG_INSTR_CLOCK_NAME;
+#endif
+ if (fast_timing && !fast_timing_used)
+ printf(_("Warning: Fast timing requested, but not available - regular timing source will be used\n"));
+ printf(_("Time source: %s\n"), time_source);
/*
* Pre-zero the statistics data structures. They're already zero by
@@ -173,8 +199,11 @@ test_timing(unsigned int duration)
total_time = duration > 0 ? duration * INT64CONST(1000000000) : 0;
- INSTR_TIME_SET_CURRENT(start_time);
- cur = INSTR_TIME_GET_NANOSEC(start_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(start_time);
+ else
+ INSTR_TIME_SET_CURRENT(start_time);
+ cur = start_time;
while (time_elapsed < total_time)
{
@@ -182,9 +211,11 @@ test_timing(unsigned int duration)
bits;
prev = cur;
- INSTR_TIME_SET_CURRENT(temp);
- cur = INSTR_TIME_GET_NANOSEC(temp);
- diff = cur - prev;
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(cur);
+ else
+ INSTR_TIME_SET_CURRENT(cur);
+ diff = INSTR_TIME_DIFF_NANOSEC(cur, prev);
/* Did time go backwards? */
if (unlikely(diff < 0))
@@ -217,11 +248,13 @@ test_timing(unsigned int duration)
largest_diff_count++;
loop_count++;
- INSTR_TIME_SUBTRACT(temp, start_time);
- time_elapsed = INSTR_TIME_GET_NANOSEC(temp);
+ time_elapsed = INSTR_TIME_DIFF_NANOSEC(cur, start_time);
}
- INSTR_TIME_SET_CURRENT(end_time);
+ if (fast_timing)
+ INSTR_TIME_SET_CURRENT_FAST(end_time);
+ else
+ INSTR_TIME_SET_CURRENT(end_time);
INSTR_TIME_SUBTRACT(end_time, start_time);
diff --git a/src/include/portability/instr_time.h b/src/include/portability/instr_time.h
index e2e339a0c4f..f02296f1026 100644
--- a/src/include/portability/instr_time.h
+++ b/src/include/portability/instr_time.h
@@ -112,10 +112,13 @@ extern int64 max_ticks_no_overflow;
*/
#if defined(__darwin__) && defined(CLOCK_MONOTONIC_RAW)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC_RAW
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC_RAW)"
#elif defined(CLOCK_MONOTONIC)
#define PG_INSTR_CLOCK CLOCK_MONOTONIC
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_MONOTONIC)"
#else
#define PG_INSTR_CLOCK CLOCK_REALTIME
+#define PG_INSTR_CLOCK_NAME "clock_gettime (CLOCK_REALTIME)"
#endif
#if defined(__x86_64__) && defined(__linux__)
@@ -174,7 +177,7 @@ pg_get_ticks(void)
}
static inline int64_t
-pg_ticks_to_ns(instr_time t)
+pg_ticks_to_ns(int64 ticks)
{
/*
* Would multiplication overflow? If so perform computation in two parts.
@@ -183,7 +186,7 @@ pg_ticks_to_ns(instr_time t)
*/
int64 ns = 0;
- if (unlikely(t.ticks > max_ticks_no_overflow))
+ if (unlikely(ticks > max_ticks_no_overflow))
{
/*
* Compute how often the maximum number of ticks fits completely into
@@ -192,7 +195,7 @@ pg_ticks_to_ns(instr_time t)
* value. In a 2nd step we adjust the number of elapsed ticks and
* convert the remaining ticks.
*/
- int64 count = t.ticks / max_ticks_no_overflow;
+ int64 count = ticks / max_ticks_no_overflow;
int64 max_ns = max_ticks_no_overflow * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
ns = max_ns * count;
@@ -201,11 +204,11 @@ pg_ticks_to_ns(instr_time t)
* Subtract the ticks that we now already accounted for, so that they
* don't get counted twice.
*/
- t.ticks -= count * max_ticks_no_overflow;
- Assert(t.ticks >= 0);
+ ticks -= count * max_ticks_no_overflow;
+ Assert(ticks >= 0);
}
- ns += t.ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
+ ns += ticks * ticks_per_ns_scaled / TICKS_TO_NS_PRECISION;
return ns;
}
@@ -226,14 +229,14 @@ pg_initialize_get_ticks()
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_get_ticks())
-#define INSTR_TIME_GET_NANOSEC(t) \
- pg_ticks_to_ns(t)
-
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ (pg_ticks_to_ns(ticks))
#else /* WIN32 */
/* Use QueryPerformanceCounter() */
+#define PG_INSTR_CLOCK_NAME "QueryPerformanceCounter"
/* helper for INSTR_TIME_SET_CURRENT / INSTR_TIME_SET_CURRENT_FAST */
static inline instr_time
@@ -265,8 +268,8 @@ GetTimerFrequency(void)
#define INSTR_TIME_SET_CURRENT(t) \
((t) = pg_query_performance_counter())
-#define INSTR_TIME_GET_NANOSEC(t) \
- ((int64) ((t).ticks * ((double) NS_PER_S / GetTimerFrequency())))
+#define INSTR_TIME_TICKS_TO_NANOSEC(ticks) \
+ ((int64) ((ticks) * ((double) NS_PER_S / GetTimerFrequency())))
#endif /* WIN32 */
@@ -285,9 +288,14 @@ GetTimerFrequency(void)
#define INSTR_TIME_SUBTRACT(x,y) \
((x).ticks -= (y).ticks)
+#define INSTR_TIME_DIFF_NANOSEC(x,y) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((x).ticks - (y).ticks))
+
#define INSTR_TIME_ACCUM_DIFF(x,y,z) \
((x).ticks += (y).ticks - (z).ticks)
+#define INSTR_TIME_GET_NANOSEC(t) \
+ (INSTR_TIME_TICKS_TO_NANOSEC((t).ticks))
#define INSTR_TIME_GET_DOUBLE(t) \
((double) INSTR_TIME_GET_NANOSEC(t) / NS_PER_S)
--
2.47.3
--vtqqtrpooseurzip--
^ permalink raw reply [nested|flat] 268+ messages in thread
end of thread, other threads:[~2025-07-27 15:48 UTC | newest]
Thread overview: 268+ messages (download: mbox mbox.gz follow: Atom feed)
-- links below jump to the message on this page --
2020-10-17 11:53 Timing of relcache inval at parallel worker init Noah Misch <[email protected]>
2020-10-20 08:35 ` Kyotaro Horiguchi <[email protected]>
2020-10-21 15:31 ` Robert Haas <[email protected]>
2020-10-24 15:29 ` Noah Misch <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
2025-07-27 15:48 [PATCH v12 3/3] pg_test_timing: Add --fast flag to test fast timing, report time source Lukas Fittl <[email protected]>
This inbox is served by agora; see mirroring instructions
for how to clone and mirror all data and code used for this inbox