[PATCH 3/6] Optimize allocations in bringetbitmap

($INBOX_DIR/description missing)  
help / color / mirror / Atom feed

[PATCH 3/6] Optimize allocations in bringetbitmap
276+ messages / 4 participants
[nested] [flat]

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 4dd29c7b10..470431010d 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------45196B023835614BDFCBD16D
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20201220.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20201220.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------310A2AE1CC4C2E2E77559E3D
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114-2.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114-2.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------310A2AE1CC4C2E2E77559E3D
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114-2.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114-2.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------FA974B75A0C3E9CA18CE7207
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210122.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210122.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------310A2AE1CC4C2E2E77559E3D
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114-2.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114-2.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------CF71AF65F7C337C37C24B045
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210112.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210112.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------CF71AF65F7C337C37C24B045
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210112.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210112.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------CF71AF65F7C337C37C24B045
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210112.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210112.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D21F32E0F9E7408E9BE1AF2E
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D21F32E0F9E7408E9BE1AF2E
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------FA974B75A0C3E9CA18CE7207
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210122.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210122.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------310A2AE1CC4C2E2E77559E3D
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114-2.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114-2.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------310A2AE1CC4C2E2E77559E3D
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114-2.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114-2.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------FA974B75A0C3E9CA18CE7207
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210122.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210122.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D21F32E0F9E7408E9BE1AF2E
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D21F32E0F9E7408E9BE1AF2E
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------CF71AF65F7C337C37C24B045
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210112.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210112.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------FA974B75A0C3E9CA18CE7207
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210122.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210122.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------FA974B75A0C3E9CA18CE7207
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210122.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210122.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------FA974B75A0C3E9CA18CE7207
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210122.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210122.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------FA974B75A0C3E9CA18CE7207
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210122.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210122.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D21F32E0F9E7408E9BE1AF2E
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D21F32E0F9E7408E9BE1AF2E
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D21F32E0F9E7408E9BE1AF2E
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------310A2AE1CC4C2E2E77559E3D
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114-2.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114-2.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D21F32E0F9E7408E9BE1AF2E
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------FA974B75A0C3E9CA18CE7207
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210122.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210122.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------CF71AF65F7C337C37C24B045
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210112.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210112.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------CF71AF65F7C337C37C24B045
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210112.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210112.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------FA974B75A0C3E9CA18CE7207
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210122.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210122.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 4dd29c7b10..470431010d 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------90362739C57C6315B5637FAD
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20201107.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20201107.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D21F32E0F9E7408E9BE1AF2E
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------310A2AE1CC4C2E2E77559E3D
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114-2.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114-2.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------310A2AE1CC4C2E2E77559E3D
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114-2.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114-2.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------CF71AF65F7C337C37C24B045
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210112.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210112.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D21F32E0F9E7408E9BE1AF2E
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------CF71AF65F7C337C37C24B045
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210112.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210112.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------310A2AE1CC4C2E2E77559E3D
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114-2.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114-2.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D21F32E0F9E7408E9BE1AF2E
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------CF71AF65F7C337C37C24B045
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210112.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210112.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D21F32E0F9E7408E9BE1AF2E
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------CF71AF65F7C337C37C24B045
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210112.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210112.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------CF71AF65F7C337C37C24B045
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210112.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210112.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D21F32E0F9E7408E9BE1AF2E
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------CF71AF65F7C337C37C24B045
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210112.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210112.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------310A2AE1CC4C2E2E77559E3D
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114-2.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114-2.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------FA974B75A0C3E9CA18CE7207
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210122.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210122.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------310A2AE1CC4C2E2E77559E3D
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114-2.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114-2.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------CF71AF65F7C337C37C24B045
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210112.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210112.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------FA974B75A0C3E9CA18CE7207
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210122.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210122.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------FA974B75A0C3E9CA18CE7207
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210122.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210122.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------CF71AF65F7C337C37C24B045
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210112.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210112.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------FA974B75A0C3E9CA18CE7207
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210122.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210122.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------FA974B75A0C3E9CA18CE7207
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210122.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210122.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------CF71AF65F7C337C37C24B045
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210112.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210112.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------310A2AE1CC4C2E2E77559E3D
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114-2.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114-2.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------FA974B75A0C3E9CA18CE7207
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210122.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210122.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------CF71AF65F7C337C37C24B045
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210112.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210112.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------CF71AF65F7C337C37C24B045
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210112.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210112.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D21F32E0F9E7408E9BE1AF2E
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------FA974B75A0C3E9CA18CE7207
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210122.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210122.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------CF71AF65F7C337C37C24B045
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210112.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210112.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------CF71AF65F7C337C37C24B045
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210112.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210112.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------310A2AE1CC4C2E2E77559E3D
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114-2.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114-2.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D21F32E0F9E7408E9BE1AF2E
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------310A2AE1CC4C2E2E77559E3D
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114-2.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114-2.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------310A2AE1CC4C2E2E77559E3D
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114-2.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114-2.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------FA974B75A0C3E9CA18CE7207
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210122.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210122.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------FA974B75A0C3E9CA18CE7207
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210122.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210122.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D21F32E0F9E7408E9BE1AF2E
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D21F32E0F9E7408E9BE1AF2E
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------310A2AE1CC4C2E2E77559E3D
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114-2.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114-2.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------310A2AE1CC4C2E2E77559E3D
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114-2.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114-2.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D21F32E0F9E7408E9BE1AF2E
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------310A2AE1CC4C2E2E77559E3D
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114-2.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114-2.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------FA974B75A0C3E9CA18CE7207
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210122.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210122.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------310A2AE1CC4C2E2E77559E3D
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114-2.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114-2.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------CF71AF65F7C337C37C24B045
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210112.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210112.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------CF71AF65F7C337C37C24B045
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210112.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210112.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D21F32E0F9E7408E9BE1AF2E
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------310A2AE1CC4C2E2E77559E3D
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114-2.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114-2.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------FA974B75A0C3E9CA18CE7207
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210122.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210122.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------FA974B75A0C3E9CA18CE7207
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210122.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210122.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D21F32E0F9E7408E9BE1AF2E
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 55851376d8..8d23f2864b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D21F32E0F9E7408E9BE1AF2E
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------0E72B707603BED22B4040825
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210211.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/8] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 4dd29c7b10..470431010d 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -372,6 +372,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -397,11 +400,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -437,9 +479,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -450,17 +492,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------291E0481B4400C0ECC8E497A
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20201108.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20201108.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..c7f7175a2a 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------22A4B241170149838D4D1F8F
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210215.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210215.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------310A2AE1CC4C2E2E77559E3D
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210114-2.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210114-2.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/9] Optimize allocations in bringetbitmap
@ 2020-09-13 10:12 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2020-09-13 10:12 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 62 +++++++++++++++++++++++++++-------
 1 file changed, 49 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 14da9ed17f..3735c41788 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -398,11 +401,50 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * Make room for per-attribute lists of scan keys that we'll pass to the
 	 * consistent support procedure. We keep null and regular keys separate,
 	 * so that we can easily pass regular keys to the consistent function.
+	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly
+	 * the same lifetime, so that's OK.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		/* regular keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		/* NULL keys */
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -438,9 +480,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -451,17 +493,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------4B194FF8F3EA3786FF9EAE1F
Content-Type: text/x-patch; charset=UTF-8;
 name="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.patch"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
 filename*0="0002-Move-IS-NOT-NULL-handling-from-BRIN-support-20210203.pa";
 filename*1="tch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 3/6] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------6A37408A2C42B5B56C3B3D49
Content-Type: text/x-patch; charset=UTF-8;
 name="0004-BRIN-bloom-indexes-20210303.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0004-BRIN-bloom-indexes-20210303.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index ce0f525c21..758c47f0f4 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------D524CD37E77B36A2883A7617
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210311.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210311.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------556A1DC61262AF15641DB204
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308b.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308b.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------82D677E9F94AC7574E460205
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210308.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210308.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 5/8] Optimize allocations in bringetbitmap
@ 2021-03-02 18:57 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-02 18:57 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index 9f2656b8d9..1f82e965f9 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -402,15 +405,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest table can have ~1600 attributes, so this may allocate a
 	 * couple kilobytes of memory). We could invent a more compact approach
 	 * (with just space for used attributes) but that would make the matching
 	 * more complicated, so it may not be a win.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/*
 	 * Preprocess the scan keys - split them into per-attribute arrays.
@@ -444,9 +484,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -457,17 +497,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.26.2


--------------AC3FFB734F56C5F52D422EFC
Content-Type: text/x-patch; charset=UTF-8;
 name="0006-BRIN-bloom-indexes-20210305.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0006-BRIN-bloom-indexes-20210305.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-21 23:24 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-21 23:24 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Reviewed-by: Alvaro Herrera <[email protected]>
Reviewed-by: Mark Dilger <[email protected]>
Reviewed-by: Alexander Korotkov <[email protected]>
Reviewed-by: Masahiko Sawada <[email protected]>
Reviewed-by: John Naylor <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index efda84aba3..0bf25fd05b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -403,15 +406,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest index can have 32 attributes, so the amount of wasted
 	 * memory is negligible. We could invent a more compact approach (with
 	 * just space for used attributes) but that would make the matching more
 	 * complex so it's not a good trade-off.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/* Preprocess the scan keys - split them into per-attribute arrays. */
 	for (keyno = 0; keyno < scan->numberOfKeys; keyno++)
@@ -443,9 +483,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -456,17 +496,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.30.2


--------------76348C5C2C314FA585652FE5
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210322.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210322.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* [PATCH 4/8] Optimize allocations in bringetbitmap
@ 2021-03-21 23:24 Tomas Vondra <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: Tomas Vondra @ 2021-03-21 23:24 UTC (permalink / raw)

The bringetbitmap function allocates memory for various purposes, which
may be quite expensive, depending on the number of scan keys. Instead of
allocating them separately, allocate one bit chunk of memory an carve it
into smaller pieces as needed - all the pieces have the same lifespan,
and it saves quite a bit of CPU and memory overhead.

Author: Tomas Vondra <[email protected]>
Reviewed-by: Alvaro Herrera <[email protected]>
Reviewed-by: Mark Dilger <[email protected]>
Reviewed-by: Alexander Korotkov <[email protected]>
Reviewed-by: Masahiko Sawada <[email protected]>
Reviewed-by: John Naylor <[email protected]>
Discussion: https://postgr.es/m/[email protected]
---
 src/backend/access/brin/brin.c | 60 ++++++++++++++++++++++++++--------
 1 file changed, 47 insertions(+), 13 deletions(-)

diff --git a/src/backend/access/brin/brin.c b/src/backend/access/brin/brin.c
index efda84aba3..0bf25fd05b 100644
--- a/src/backend/access/brin/brin.c
+++ b/src/backend/access/brin/brin.c
@@ -373,6 +373,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	int		   *nkeys,
 			   *nnullkeys;
 	int			keyno;
+	char	   *ptr;
+	Size		len;
+	char	   *tmp PG_USED_FOR_ASSERTS_ONLY;
 
 	opaque = (BrinOpaque *) scan->opaque;
 	bdesc = opaque->bo_bdesc;
@@ -403,15 +406,52 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 	 * We keep null and regular keys separate, so that we can pass just the
 	 * regular keys to the consistent function easily.
 	 *
+	 * To reduce the allocation overhead, we allocate one big chunk and then
+	 * carve it into smaller arrays ourselves. All the pieces have exactly the
+	 * same lifetime, so that's OK.
+	 *
 	 * XXX The widest index can have 32 attributes, so the amount of wasted
 	 * memory is negligible. We could invent a more compact approach (with
 	 * just space for used attributes) but that would make the matching more
 	 * complex so it's not a good trade-off.
 	 */
-	keys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nullkeys = palloc0(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
-	nkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
-	nnullkeys = palloc0(sizeof(int) * bdesc->bd_tupdesc->natts);
+	len =
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* regular keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts) +
+		MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts) +	/* NULL keys */
+		MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys) * bdesc->bd_tupdesc->natts +
+		MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	ptr = palloc(len);
+	tmp = ptr;
+
+	keys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nullkeys = (ScanKey **) ptr;
+	ptr += MAXALIGN(sizeof(ScanKey *) * bdesc->bd_tupdesc->natts);
+
+	nkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	nnullkeys = (int *) ptr;
+	ptr += MAXALIGN(sizeof(int) * bdesc->bd_tupdesc->natts);
+
+	for (int i = 0; i < bdesc->bd_tupdesc->natts; i++)
+	{
+		keys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+
+		nullkeys[i] = (ScanKey *) ptr;
+		ptr += MAXALIGN(sizeof(ScanKey) * scan->numberOfKeys);
+	}
+
+	Assert(tmp + len == ptr);
+
+	/* zero the number of keys */
+	memset(nkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
+	memset(nnullkeys, 0, sizeof(int) * bdesc->bd_tupdesc->natts);
 
 	/* Preprocess the scan keys - split them into per-attribute arrays. */
 	for (keyno = 0; keyno < scan->numberOfKeys; keyno++)
@@ -443,9 +483,9 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		{
 			FmgrInfo   *tmp;
 
-			/* No key/null arrays for this attribute. */
-			Assert((keys[keyattno - 1] == NULL) && (nkeys[keyattno - 1] == 0));
-			Assert((nullkeys[keyattno - 1] == NULL) && (nnullkeys[keyattno - 1] == 0));
+			/* First time we see this attribute, so no key/null keys. */
+			Assert(nkeys[keyattno - 1] == 0);
+			Assert(nnullkeys[keyattno - 1] == 0);
 
 			tmp = index_getprocinfo(idxRel, keyattno,
 									BRIN_PROCNUM_CONSISTENT);
@@ -456,17 +496,11 @@ bringetbitmap(IndexScanDesc scan, TIDBitmap *tbm)
 		/* Add key to the proper per-attribute array. */
 		if (key->sk_flags & SK_ISNULL)
 		{
-			if (!nullkeys[keyattno - 1])
-				nullkeys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			nullkeys[keyattno - 1][nnullkeys[keyattno - 1]] = key;
 			nnullkeys[keyattno - 1]++;
 		}
 		else
 		{
-			if (!keys[keyattno - 1])
-				keys[keyattno - 1] = palloc0(sizeof(ScanKey) * scan->numberOfKeys);
-
 			keys[keyattno - 1][nkeys[keyattno - 1]] = key;
 			nkeys[keyattno - 1]++;
 		}
-- 
2.30.2


--------------76348C5C2C314FA585652FE5
Content-Type: text/x-patch; charset=UTF-8;
 name="0005-BRIN-bloom-indexes-20210322.patch"
Content-Transfer-Encoding: 8bit
Content-Disposition: attachment;
 filename="0005-BRIN-bloom-indexes-20210322.patch"



^ permalink  raw  reply  [nested|flat] 276+ messages in thread

* Re: Commitfest manager November 2023
@ 2023-11-02 09:41 John Naylor <[email protected]>
  0 siblings, 0 replies; 276+ messages in thread

From: John Naylor @ 2023-11-02 09:41 UTC (permalink / raw)
  To: [email protected]; +Cc: pgsql-hackers

On Thu, Nov 2, 2023 at 4:35 PM Daniel Gustafsson <[email protected]> wrote:
>
> > On 1 Nov 2023, at 07:33, John Naylor <[email protected]> wrote:
> >
> > I didn't see any recent mentions in the archives, so I'll volunteer to
> > be CF manager for 2023-11.
>
> You probably need some extra admin privileges on your account for accessing the
> CFM functionality, in the meantime I've switched the 202311 CF to InProgress
> and marked 202401 as Open.

Thanks for taking care of that!

(Per the wiki, I requested admin privs on pgsql-www, so still awaiting that)






^ permalink  raw  reply  [nested|flat] 276+ messages in thread

end of thread, other threads:[~2023-11-02 09:41 UTC | newest]

Thread overview: 276+ messages (download: mbox mbox.gz follow: Atom feed)
-- links below jump to the message on this page --
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2020-09-13 10:12 [PATCH 3/9] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 3/6] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-02 18:57 [PATCH 5/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-21 23:24 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2021-03-21 23:24 [PATCH 4/8] Optimize allocations in bringetbitmap Tomas Vondra <[email protected]>
2023-11-02 09:41 Re: Commitfest manager November 2023 John Naylor <[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