Cope with inplace update making catcache stale during TOAST fetch.
authorNoah Misch <[email protected]>
Fri, 28 Jun 2024 02:21:06 +0000 (19:21 -0700)
committerNoah Misch <[email protected]>
Fri, 28 Jun 2024 02:21:13 +0000 (19:21 -0700)
commit11f3815d6af840dc9d557cc680ec477931a6825b
tree3007444b0231211ed637ee0fdc4d516b75a62d73
parent7af775bbae5724272f3b72766df977f07fa5455a
Cope with inplace update making catcache stale during TOAST fetch.

This extends ad98fb14226ae6456fbaed7990ee7591cbe5efd2 to invals of
inplace updates.  Trouble requires an inplace update of a catalog having
a TOAST table, so only pg_database was at risk.  (The other catalog on
which core code performs inplace updates, pg_class, has no TOAST table.)
Trouble would require something like the inplace-inval.spec test.
Consider GRANT ... ON DATABASE fetching a stale row from cache and
discarding a datfrozenxid update that vac_truncate_clog() has already
relied upon.  Back-patch to v12 (all supported versions).

Reviewed (in an earlier version) by Robert Haas.

Discussion: https://postgr.es/m/20240114201411[email protected]
Discussion: https://postgr.es/m/20240512232923[email protected]
src/backend/catalog/catalog.c
src/backend/utils/cache/catcache.c
src/include/catalog/catalog.h