Horton

Re: BUG #19508: pg_buffercache_pages() crashes the backend with an incompatible caller-supplied record definition

Ayush Tiwari <ayushtiwari.slg01@gmail.com>

View thread View on postgr.es
From: Ayush Tiwari <ayushtiwari.slg01@gmail.com>
To: Fujii Masao <masao.fujii@gmail.com>
Cc: Калинин Никита <n.kalinin@postgrespro.ru>, PostgreSQL mailing lists <pgsql-bugs@lists.postgresql.org>
Date: 2026-06-05T08:47:58Z
Lists: pgsql-bugs 
Hi,

On Fri, 5 Jun, 2026, 13:28 Fujii Masao, <masao.fujii@gmail.com> wrote:

> On Fri, Jun 5, 2026 at 12:42 PM Ayush Tiwari
> <ayushtiwari.slg01@gmail.com> wrote:
> > One small nit: build_buffercache_pages_tupledesc() names attribute 8
> > "usage_count", while the existing pg_buffercache view and the test use
> > "usagecount". This probably does not affect the tupledesc_match() check,
> > but I think it would be better to keep the existing spelling for
> > consistency.
>
> Agreed. I've fixed that and attached an updated version of the patch.
>

LGTM.

Regards,
Ayush

>

Commits

  1. pg_buffercache: restore rowtype verification in pg_buffercache_pages()

  2. Modernize and optimize pg_buffercache_pages()