Re: Inconsistent update in the MERGE command

From: Dean Rasheed <dean(dot)a(dot)rasheed(at)gmail(dot)com>
To: Yugo Nagata <nagata(at)sraoss(dot)co(dot)jp>
Cc: Dmitry <dsy(dot)075(at)yandex(dot)ru>, pgsql-hackers(at)lists(dot)postgresql(dot)org
Subject: Re: Inconsistent update in the MERGE command
Date: 2025-09-05 07:47:19
Message-ID: CAEZATCVhtMEW+rq4gdWcxyFjOfZGOwc+8tSp4ZxKbvP+eW2SWQ@mail.gmail.com
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On Thu, 4 Sept 2025 at 16:03, Yugo Nagata <nagata(at)sraoss(dot)co(dot)jp> wrote:
>
> > I've updated that to use tupleid in the attached v3 patch, and added a
> > couple more isolation tests. In practice, however, I don't think that
> > error can ever happen because this check follows table_tuple_lock()
> > which has a similar test which will always error out first. I was
> > tempted to simply remove this check from ExecMergeMatched(), but I
> > think perhaps it's worth keeping just in case.
>
> ItemPointerIndicatesMovedPartitions() is checked in heapam_tuple_lock(),
> but other table access methods might not perform this check (though I'm
> not sure if this is acceptable), so it may still make sense to keep it.

Good point.

> > Also, I thought that it's worth updating the comments for
> > table_tuple_lock() and TM_FailureData to make it clearer that it
> > updates its tid argument, and which fields of TM_FailureData can be
> > relied upon in the different cases.
>
> +1
> This should help prevent misuse of the argument in the future.

OK, thanks. Pushed.

In v15 and v16 the code was a little different, and it was actually
much more obvious that it thought that table_tuple_lock() didn't
update the tupleid passed to it, because it was explicitly updating it
from tmfd->ctid.

It was also slightly harder to trigger the error in v15 and v16, so I
had to tweak the tests a little. In doing so, I realised that it's not
actually necessary to have 3 sessions to reproduce this error -- it
only requires 2 sessions, as long as the "other" session does 2
updates. So I did it that way in all branches, which simplified them a
bit.

Regards,
Dean

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Dmitry Mityugov 2025-09-05 08:11:42 Re: --with-llvm on 32-bit platforms?
Previous Message Richard Guo 2025-09-05 07:35:17 Re: Eager aggregation, take 3