Re: [HACKERS] Question on win32 semaphore simulation

Bruce Momjian <pgman(at)candle(dot)pha(dot)pa(dot)us> writes:
> Peter Brant wrote:
>> We've never been able reproduce (or even trigger) the original "sem_ctl
>> failed" error in a testing environment so it would be hard to say if the
>> changes to win32/sema.c have an impact on it or not. On the other hand,
>> win32_sema.c seems to solve the pgbench lockups reported earlier by Jim
>> N. and it successfully completes a reasonably brutal stress test with
>> real world data and real world queries (which at least is a good
>> indication that it basically works).

> OK, let's consider the item closed. We didn't backpatch the new
> win32_sema.c file to 8.1.X or 8.0.X, so let'see if we get more reports.

Based on that, backpatching the new win32_sema.c implementation is
probably more defensible than applying the proposed smaller patch
anyway; it's survived more testing.

My inclination is to do nothing to the back branches, but if we get more
field reports of trouble with them, maybe that's what to do. (I'd be
happier if 8.2 gets through beta first, as I'm still a bit worried about
the do-all-Windows-versions-act-the-same bit.)

regards, tom lane