Horton

Re: Server crash (FailedAssertion) due to catcache refcount mis-handling

Tom Lane <tgl@sss.pgh.pa.us>

View thread View on postgr.es
From: Tom Lane <tgl@sss.pgh.pa.us>
To: Robert Haas <robertmhaas@gmail.com>
Cc: Jeevan Chalke <jeevan.chalke@enterprisedb.com>, PostgreSQL Hackers <pgsql-hackers@postgresql.org>
Date: 2017-08-13T19:05:13Z
Lists: pgsql-hackers 
Robert Haas <robertmhaas@gmail.com> writes:
> On Tue, Aug 8, 2017 at 10:48 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
>> Robert Haas <robertmhaas@gmail.com> writes:
>>> On Tue, Aug 8, 2017 at 4:34 PM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
>>>> In the meantime, I think my vote would be to remove AtEOXact_CatCache.

>>> In all supported branches?

>> Whatever we do about this issue, I don't feel a need to do it further
>> back than HEAD.  It's a non-problem except in an assert-enabled build,
>> and we don't recommend running those for production, only development.

> Sure, but people still do testing and development against older
> branches - bug fixes, for example.  It doesn't make much sense to me
> to leave code that we know does the wrong thing in the back branches.

Not having heard anyone arguing against that, I'll go make it so,
ie AtEOXact_CatCache is toast in all branches.

			regards, tom lane

Commits

  1. Handle elog(FATAL) during ROLLBACK more robustly.

  2. Remove AtEOXact_CatCache().