Horton

Re: [CLOBBER_CACHE]Server crashed with segfault 11 while executing clusterdb

Michael Paquier <michael@paquier.xyz>

View thread View on postgr.es
From: Michael Paquier <michael@paquier.xyz>
To: Alvaro Herrera <alvherre@alvh.no-ip.org>
Cc: Amul Sul <sulamul@gmail.com>, Kyotaro Horiguchi <horikyota.ntt@gmail.com>, Tom Lane <tgl@sss.pgh.pa.us>, Noah Misch <noah@leadboat.com>, Neha Sharma <neha.sharma@enterprisedb.com>, Amit Langote <amitlangote09@gmail.com>, PostgreSQL-development <pgsql-hackers@postgresql.org>
Date: 2021-04-19T06:55:08Z
Lists: pgsql-hackers 
On Fri, Apr 09, 2021 at 06:45:45PM -0400, Alvaro Herrera wrote:
> We forgot this patch earlier in the commitfest.  Do people think we
> should still get it in on this cycle?  I'm +1 on that, since it's a
> safety feature poised to prevent more bugs than it's likely to
> introduce.

No objections from here to do that now even after feature freeze.  I
also wonder, while looking at that, why you don't just remove the last
call within src/backend/catalog/heap.c.  This way, nobody is tempted
to use RelationOpenSmgr() anymore, and it could just be removed from
rel.h.
--
Michael

Commits

  1. Replace RelationOpenSmgr() with RelationGetSmgr().

  2. Avoid possible crash while finishing up a heap rewrite.

  3. Redefine pg_class.reltuples to be -1 before the first VACUUM or ANALYZE.