pgsql: Replace ia64 S_UNLOCK compiler barrier with a full memory barrie

From: Andres Freund <andres(at)anarazel(dot)de>
To: pgsql-committers(at)postgresql(dot)org
Subject: pgsql: Replace ia64 S_UNLOCK compiler barrier with a full memory barrie
Date: 2015-06-29 12:54:07
Message-ID: E1Z9YZb-0005iK-K3@gemulon.postgresql.org
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-committers

Replace ia64 S_UNLOCK compiler barrier with a full memory barrier.

_Asm_sched_fence() is just a compiler barrier, not a memory barrier. But
spinlock release on IA64 needs, at the very least, release
semantics. Use a full barrier instead.

This might be the cause for the occasional failures on buildfarm member
anole.

Discussion: 20150629101108(dot)GB17640(at)alap3(dot)anarazel(dot)de

Branch
------
master

Details
-------
http://git.postgresql.org/pg/commitdiff/07cb8b02ab4c8b65bb2e3b87ad2402fdc6cce978

Modified Files
--------------
src/include/storage/s_lock.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

Browse pgsql-committers by date

  From Date Subject
Next Message Andrew Dunstan 2015-06-29 14:02:32 Re: pgsql: Run the C portions of guc-file.l through pgindent.
Previous Message Andres Freund 2015-06-29 10:02:54 Re: pgsql: Fix the fallback memory barrier implementation to be reentrant.