Re: BUG #18146: Rows reappearing in Tables after Auto-Vacuum Failure in PostgreSQL on Windows

On Tue, Sep 10, 2024 at 12:43 PM Thomas Munro <thomas.munro@gmail.com> wrote:
> It is possible that an extension that messes with smgrsw[] would not
> like this in a minor release:
>
> -               smgrsw[reln->smgr_which].smgr_truncate(reln,
> forknum[i], nblocks[i]);
> +               smgrsw[reln->smgr_which].smgr_truncate(reln, forknum[i],
> +
>                     old_nblocks[i], nblocks[i]);

Scratch that concern, smgrsw is static in smgr.c for now (h/t to
Andres for pointing that out in an off-list chat when I described
this...).  I must have confused myself with proposals I've read to
make it so an extension *could* mess with this stuff, but it's not
done yet.

Also if this goes somewhere I should note in the commit message that
the analysis of the WaitIO issue was from Alexander L in bug #18426.