Tom Lane dab421d2f0 Seems I was too optimistic in supposing that sinval's maxMsgNum could be ... read and written without a lock. The value itself is atomic, sure, but on processors with weak memory ordering it's possible for a reader to see the value change before it sees the associated message written into the buffer array. Fix by introducing a spinlock that's used just to read and write maxMsgNum. (We could do this with less overhead if we recognized a concept of "memory access barrier"; is it worth introducing such a thing? At the moment probably not --- I can't measure any clear slowdown from adding the spinlock, so this solution is probably fine.) Per buildfarm results.		2008-06-20 00:24:53 +00:00
..
access	Improve our #include situation by moving pointer types away from the	2008-06-19 00:46:06 +00:00
bootstrap	Restructure some header files a bit, in particular heapam.h, by removing some	2008-05-12 00:00:54 +00:00
catalog	Improve our #include situation by moving pointer types away from the	2008-06-19 00:46:06 +00:00
commands	Improve our #include situation by moving pointer types away from the	2008-06-19 00:46:06 +00:00
executor	Improve our #include situation by moving pointer types away from the	2008-06-19 00:46:06 +00:00
lib	Refactor backend makefiles to remove lots of duplicate code	2008-02-19 10:30:09 +00:00
libpq	Use error message wordings for permissions checks on .pgpass and SSL private	2008-03-31 02:43:14 +00:00
main	Refactor backend makefiles to remove lots of duplicate code	2008-02-19 10:30:09 +00:00
nodes	Improve our #include situation by moving pointer types away from the	2008-06-19 00:46:06 +00:00
optimizer	Improve our #include situation by moving pointer types away from the	2008-06-19 00:46:06 +00:00
parser	Improve our #include situation by moving pointer types away from the	2008-06-19 00:46:06 +00:00
po	Translation updates	2008-01-31 18:04:52 +00:00
port	Add $PostgreSQL$ markers to a lot of files that were missing them.	2008-05-17 01:28:26 +00:00
postmaster	Rewrite the sinval messaging mechanism to reduce contention and avoid	2008-06-19 21:32:56 +00:00
regex	Refactor backend makefiles to remove lots of duplicate code	2008-02-19 10:30:09 +00:00
rewrite	Improve our #include situation by moving pointer types away from the	2008-06-19 00:46:06 +00:00
snowball	Implement a few changes to how shared libraries and dynamically loadable	2008-04-07 14:15:58 +00:00
storage	Seems I was too optimistic in supposing that sinval's maxMsgNum could be	2008-06-20 00:24:53 +00:00
tcop	Rearrange ALTER TABLE syntax processing as per my recent proposal: the	2008-06-15 01:25:54 +00:00
tsearch	Fix a few places that were non-multibyte-safe in tsearch configuration file	2008-06-19 16:52:24 +00:00
utils	Rewrite the sinval messaging mechanism to reduce contention and avoid	2008-06-19 21:32:56 +00:00
common.mk	We need to rebuild objfiles.txt when one of the subdirectories' objfiles.txt	2008-03-17 18:24:56 +00:00
Makefile	Enable probes to work with Mac OS X Leopard and other OSes that will	2008-03-17 19:44:41 +00:00
nls.mk	Add a new ereport auxiliary function errdetail_log(), which works the same as	2008-03-24 18:08:47 +00:00