pgsql: Mark pageinspect's disk-accessing functions as parallel restrict - Mailing list pgsql-committers

From Tom Lane
Subject pgsql: Mark pageinspect's disk-accessing functions as parallel restrict
Date
Msg-id E1oxDXh-000GDs-V1@gemulon.postgresql.org
Whole thread Raw
List pgsql-committers
Mark pageinspect's disk-accessing functions as parallel restricted.

These functions have been marked parallel safe, but the buildfarm's
response to commit e2933a6e1 exposed the flaw in that thinking:
if you try to use them on a temporary table, and they run inside
a parallel worker, they'll fail with "cannot access temporary tables
during a parallel operation".

Fix that by marking them parallel restricted instead.  Maybe someday
we'll have a better answer and can reverse this decision.

Back-patch to v15.  To go back further, we'd have to devise variant
versions of pre-1.10 pageinspect versions.  Given the lack of field
complaints, it doesn't seem worth the trouble.  We'll just deem
this case unsupported pre-v15.  (If anyone does complain, it might
be good enough to update the markings manually in their DBs.)

Discussion: https://postgr.es/m/E1ox94a-000EHu-VH@gemulon.postgresql.org

Branch
------
master

Details
-------
https://git.postgresql.org/pg/commitdiff/aeaaf520f409cf314f97c811d2713c99858f035d

Modified Files
--------------
contrib/pageinspect/Makefile                    |  3 ++-
contrib/pageinspect/meson.build                 |  1 +
contrib/pageinspect/pageinspect--1.10--1.11.sql | 28 +++++++++++++++++++++++++
contrib/pageinspect/pageinspect.control         |  2 +-
4 files changed, 32 insertions(+), 2 deletions(-)


pgsql-committers by date:

Previous
From: Tom Lane
Date:
Subject: Re: pgsql: Prevent instability in contrib/pageinspect's regression test.
Next
From: Tom Lane
Date:
Subject: pgsql: Revert "Prevent instability in contrib/pageinspect's regression