Re: autovacuum launcher crash: assert in pgstat_count_io_op (IOOP_EXTEND on pg_database's VM)

From: Melanie Plageman <melanieplageman(at)gmail(dot)com>
To: Ewan Young <kdbase(dot)hack(at)gmail(dot)com>
Cc: pgsql-hackers(at)lists(dot)postgresql(dot)org
Subject: Re: autovacuum launcher crash: assert in pgstat_count_io_op (IOOP_EXTEND on pg_database's VM)
Date: 2026-06-24 18:59:22
Message-ID: CAAKRu_abOOp8yMi5czfEmtANOFbbS-2TQ1VtyQXi8iKiEm48Uw@mail.gmail.com
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Tue, Jun 2, 2026 at 10:03 AM Ewan Young <kdbase(dot)hack(at)gmail(dot)com> wrote:
>
> That all makes sense — your plan sounds right to me. Happy to go with
> relaxing pgstat_tracks_io_op() for 19 and leaving the
> table_beginscan_catalog() flags work for 20, and to drop my patch.
>
> Reproducer attached.
>
> Thanks so much for the quick and thoughtful response — really appreciate it!

Thanks for the report and reproducer! The fix for 19 we discussed is
now committed in 4abf411e232 (and I've marked the open item closed).
I've added a note to my personal to-do for 20 to enable setting the VM
on-access for catalog tables, but feel free to propose a patch if
you'd like.

- Melanie

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Tristan Partin 2026-06-24 19:13:36 Re: Add MIN/MAX aggregate support for uuid
Previous Message Marcos Pegoraro 2026-06-24 18:31:03 Re: Get rid of "Section.N.N.N" on DOCs