Zeugswetter Andreas SB SD wrote:
> I don't have AIX 5 here, so cannot test, sorry.
> But yes, it should definitely work.
>
> There is a known possible performance improvement for concurrent sessions on
> multiprocessor AIX machines. The now depricated "cs(3)" used for the AIX TAS
> implementation should be replaced with fetch_and_or, compare_and_swap or the
> like. I just haven't got round to doing a patch that I trust.
>
> Maybe Tatsuo can say more ?
cs() is gone in 7.3beta because cs() was failing on SMP machines anyway.
The new code is:
#define TAS(lock) _check_lock(lock, 0, 1)
#define S_UNLOCK(lock) _clear_lock(lock, 0)
--
Bruce Momjian | http://candle.pha.pa.us
[EMAIL PROTECTED] | (610) 359-1001
+ If your life is a hard drive, | 13 Roberts Road
+ Christ can be your backup. | Newtown Square, Pennsylvania 19073
---------------------------(end of broadcast)---------------------------
TIP 6: Have you searched our list archives?
http://archives.postgresql.org
