Re: PG18: use-after-free in exec partition pruning after an EPQ recheck in LockRows

From: Andrew Dunstan <andrew(at)dunslane(dot)net>
To: David Rowley <dgrowleyml(at)gmail(dot)com>, Andrey Rachitskiy <pl0h0yp1(at)gmail(dot)com>
Cc: Vladimir Savin <vladimir(at)encord(dot)com>, pgsql-bugs(at)lists(dot)postgresql(dot)org
Subject: Re: PG18: use-after-free in exec partition pruning after an EPQ recheck in LockRows
Date: 2026-10-09 14:28:31
Message-ID: ae28b629-c48d-4752-aa89-d4cb3121ed08@dunslane.net
Views: Whole Thread | Raw Message | Download mbox | Resend email
Thread:
Lists: pgsql-bugs


On 2026-10-08 Th 6:04 PM, David Rowley wrote:
> On Thu, 8 Oct 2026 at 21:30, Andrey Rachitskiy <pl0h0yp1(at)gmail(dot)com> wrote:
>> Agreed. Skipping on es_epq_active was answering the wrong question.
>> The invariant is that InitExecPartitionPruneContexts() must not run
>> twice on a shared PartitionPruneState.
>> Your initialized flag is the right fix.
> Pushed.
>

per buildfarm, this is an ABI break on release 18, apparently needs an
update to .abi-compliance-history

cheers

andrew

--
Andrew Dunstan
EDB: https://www.enterprisedb.com

In response to

Browse pgsql-bugs by date

  From Date Subject
Next Message Chinmay Kanchi 2026-10-09 17:04:44 GIN disagrees with seq scan when NOT and AND operators are combined
Previous Message Matthias van de Meent 2026-10-09 12:56:09 Re: BUG #19519: REPACK can fail due to missing chunk for toast value