From b70d4a62ee412d09c39428f6112d54d6290fa3d5 Mon Sep 17 00:00:00 2001
From: Tom Lane <tgl@sss.pgh.pa.us>
Date: Sun, 12 Aug 2007 20:39:14 +0000
Subject: [PATCH] Remove an "optimization" I installed in 2001, to make
repalloc() attempt to enlarge the memory chunk in-place when it was feasible
to do so. This turns out to not work well at all for scenarios involving
repeated cycles of palloc/repalloc/pfree: the eventually freed chunks go into
the wrong freelist for the next initial palloc request, and so we consume
memory indefinitely. While that could be defended against, the number of
cases where the optimization can still be applied drops significantly, and
adjusting the initial sizes of StringInfo buffers makes it drop to almost
nothing. Seems better to just remove the extra complexity. Per recent
discussion and testing.
---
src/backend/utils/mmgr/aset.c | 59 ++++++-----------------------------
1 file changed, 10 insertions(+), 49 deletions(-)
diff --git a/src/backend/utils/mmgr/aset.c b/src/backend/utils/mmgr/aset.c
index f517e05bc9c..7d080af86de 100644
--- a/src/backend/utils/mmgr/aset.c
+++ b/src/backend/utils/mmgr/aset.c
@@ -11,7 +11,7 @@
* Portions Copyright (c) 1994, Regents of the University of California
*
* IDENTIFICATION
- * $PostgreSQL: pgsql/src/backend/utils/mmgr/aset.c,v 1.73 2007/08/07 06:25:14 neilc Exp $
+ * $PostgreSQL: pgsql/src/backend/utils/mmgr/aset.c,v 1.74 2007/08/12 20:39:14 tgl Exp $
*
* NOTE:
* This is a new (Feb. 05, 1999) implementation of the allocation set
@@ -932,57 +932,18 @@ AllocSetRealloc(MemoryContext context, void *pointer, Size size)
else
{
/*
- * Small-chunk case. If the chunk is the last one in its block, there
- * might be enough free space after it that we can just enlarge the
- * chunk in-place. It's relatively painful to find the containing
- * block in the general case, but we can detect last-ness quite
- * cheaply for the typical case where the chunk is in the active
- * (topmost) allocation block. (At least with the regression tests
- * and code as of 1/2001, realloc'ing the last chunk of a non-topmost
- * block hardly ever happens, so it's not worth scanning the block
- * list to catch that case.)
- *
- * NOTE: must be careful not to create a chunk of a size that
- * AllocSetAlloc would not create, else we'll get confused later.
+ * Small-chunk case. We just do this by brute force, ie, allocate a
+ * new chunk and copy the data. Since we know the existing data isn't
+ * huge, this won't involve any great memcpy expense, so it's not
+ * worth being smarter. (At one time we tried to avoid memcpy when
+ * it was possible to enlarge the chunk in-place, but that turns out
+ * to misbehave unpleasantly for repeated cycles of
+ * palloc/repalloc/pfree: the eventually freed chunks go into the
+ * wrong freelist for the next initial palloc request, and so we leak
+ * memory indefinitely. See pgsql-hackers archives for 2007-08-11.)
*/
AllocPointer newPointer;
- if (size <= set->allocChunkLimit)
- {
- AllocBlock block = set->blocks;
- char *chunk_end;
-
- chunk_end = (char *) chunk + (oldsize + ALLOC_CHUNKHDRSZ);
- if (chunk_end == block->freeptr)
- {
- /* OK, it's last in block ... is there room? */
- Size freespace = block->endptr - block->freeptr;
- int fidx;
- Size newsize;
- Size delta;
-
- fidx = AllocSetFreeIndex(size);
- newsize = 1 << (fidx + ALLOC_MINBITS);
- Assert(newsize >= oldsize);
- delta = newsize - oldsize;
- if (freespace >= delta)
- {
- /* Yes, so just enlarge the chunk. */
- block->freeptr += delta;
- chunk->size += delta;
-#ifdef MEMORY_CONTEXT_CHECKING
- chunk->requested_size = size;
- /* set mark to catch clobber of "unused" space */
- if (size < chunk->size)
- ((char *) pointer)[size] = 0x7E;
-#endif
- return pointer;
- }
- }
- }
-
- /* Normal small-chunk case: just do it by brute force. */
-
/* allocate new chunk */
newPointer = AllocSetAlloc((MemoryContext) set, size);
--
GitLab