postgresql/src/backend/access
Tom Lane 5c8eb929e6 When telling the bgwriter that we need a checkpoint because too much xlog
has been consumed, recheck against the latest value of RedoRecPtr before
really sending the signal.  This avoids useless checkpoint activity if
XLogWrite is executed when we have a very stale local copy of RedoRecPtr.
The potential for useless checkpoint is very much worse in 8.3 because of
the walwriter process (which never does XLogInsert), so while this behavior
was intentional, it needs to be changed.  Per report from Itagaki Takahiro.
2007-10-12 19:39:59 +00:00
..
common Clarify some error messages about duplicate things. 2007-06-03 22:16:03 +00:00
gin HOT updates. When we update a tuple without changing any of its indexed 2007-09-20 17:56:33 +00:00
gist HOT updates. When we update a tuple without changing any of its indexed 2007-09-20 17:56:33 +00:00
hash Fix regex, LIKE, and some other second-rank text-manipulation functions 2007-09-21 22:52:52 +00:00
heap Remove incorrect use of VARSIZE() on a toasted datum. We can just remove it 2007-10-11 18:19:58 +00:00
index HOT updates. When we update a tuple without changing any of its indexed 2007-09-20 17:56:33 +00:00
nbtree HOT updates. When we update a tuple without changing any of its indexed 2007-09-20 17:56:33 +00:00
transam When telling the bgwriter that we need a checkpoint because too much xlog 2007-10-12 19:39:59 +00:00
Makefile Replace useless uses of := by = in makefiles. 2007-02-09 15:56:00 +00:00