| From: | tgl(at)svr1(dot)postgresql(dot)org (Tom Lane) |
|---|---|
| To: | pgsql-committers(at)postgresql(dot)org |
| Subject: | pgsql: Fix race condition in multixact code: it's possible to try to |
| Date: | 2005-10-28 17:27:30 |
| Message-ID: | 20051028172730.10940DB186@svr1.postgresql.org |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-committers pgsql-patches |
Log Message:
-----------
Fix race condition in multixact code: it's possible to try to read a
multixact's starting offset before the offset has been stored into the
SLRU file. A simple fix would be to hold the MultiXactGenLock until the
offset has been stored, but that looks like a big concurrency hit. Instead
rely on knowledge that unset offsets will be zero, and loop when we see
a zero. This requires a little extra hacking to ensure that zero is never
a valid value for the offset. Problem reported by Matteo Beccati, fix
ideas from Martijn van Oosterhout, Alvaro Herrera, and Tom Lane.
Modified Files:
--------------
pgsql/src/backend/access/transam:
multixact.c (r1.9 -> r1.10)
(http://developer.postgresql.org/cvsweb.cgi/pgsql/src/backend/access/transam/multixact.c.diff?r1=1.9&r2=1.10)
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Tom Lane | 2005-10-28 17:32:23 | pgsql: Clean up AIX build to avoid 'duplicate symbol' warnings, by |
| Previous Message | Teodor Sigaev | 2005-10-28 13:05:07 | pgsql: 1 Fix problem with lost precision in rank with OR-ed lexemes 2 |
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Bruce Momjian | 2005-10-28 17:48:42 | Re: [COMMITTERS] pgsql: Fix race condition in multixact code: it's possible |
| Previous Message | Tom Lane | 2005-10-28 16:50:01 | Re: [PATCHES] TODO Item - Add system view to show free space map |