On Mon, 2023-01-09 at 14:51 -0800, Nathan Bossart wrote:
> On Tue, Jan 03, 2023 at 03:45:49PM -0800, Nathan Bossart wrote:
> > I'd like to get this fixed, but I have yet to hear thoughts on the
> > suggested approach. I'll proceed with adjusting the tests and
> > documentation shortly unless someone objects.
>
> As promised, here is a new version of the patch with adjusted tests
> and
> documentation.
The current patch doesn't handle the case properly where you have
partitions that have toast tables. An easy fix by recursing, but I
think we might want to do things differently anyway.
I'm hesitant to add an index to pg_class just for the privilege checks
on toast tables, and I don't think we need to. Instead, we can just
skip the privilege check on a toast table if it's not referenced
directly, because we already checked the privileges on the parent, and
we still hold the session lock so nothing strange should have happened.
I'll look into that, but so far it looks like it should come out
cleanly for toast tables. The way you're checking privileges on the
partitioned tables is fine.
--
Jeff Davis
PostgreSQL Contributor Team - AWS