Feedback-ID: i0fe9450f:Fastmail
Date: Mon, 19 Jan 2026 17:14:46 +0900
From: Michael Paquier <michael@paquier.xyz>
To: Sami Imseih <samimseih@gmail.com>
Cc: Dmitry Dolgov <9erthalion6@gmail.com>,
	PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>
Subject: Re: [[BUG] pg_stat_statements crashes with var and non-var
 expressions in IN clause
Message-ID: <aW3n9ocMo_SH0TQJ@paquier.xyz>
References: <CAA5RZ0ts9qiONnHjjHxPxtePs22GBo4d3jZ_s2BQC59AN7XbAA@mail.gmail.com>
 <nekbxv5payvd26jf6q6b57w732cmct7ralb7ajwayr6xggifws@wc7yiimdoylk>
 <CAA5RZ0vNMm1v8ai_5zid5M8uf8Z7V7TH7Bh8B5ZHKbeLieU=gQ@mail.gmail.com>
 <w4olj5ddpehhipzggtdmhm23uermzugudyogq462lcytnrmgwj@md4l5p62zrt6>
 <CAA5RZ0vpwnB+Bmm7spRF-z2xfXhVq94-LHoDtno+_ok=gbThAw@mail.gmail.com>
MIME-Version: 1.0
Content-Type: multipart/signed; micalg=pgp-sha512;
	protocol="application/pgp-signature"; boundary="Pygv+12Q3n0rK3+G"
Content-Disposition: inline
In-Reply-To: <CAA5RZ0vpwnB+Bmm7spRF-z2xfXhVq94-LHoDtno+_ok=gbThAw@mail.gmail.com>
Archived-At: <https://www.postgresql.org/message-id/aW3n9ocMo_SH0TQJ%40paquier.xyz>
Precedence: bulk


--Pygv+12Q3n0rK3+G
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Thu, Jan 15, 2026 at 02:53:20PM -0600, Sami Imseih wrote:
>> As far as I recall it was an
>> intentional design decision, and to adjust newa->list_end in such
>> scenarios we need to solve the original problem of finding an end
>> location of any arbitrary expression,
>=20
> I'm not sure I follow. How would we adjust newa->list_end?

I may be missing something, of course, but Dmitry's message reads as
follows to me:
"It was a design choice to track the start and end locations because
we needed that for arrays, and resetting the locations if we have Vars
in the list like you are suggesting is OK by me."

I am going to re-read a bit more the area and think about a couple of
patterns to test, but I think that your suggested solution should be
OK to disable the list squashing as you are suggesting in this case:
there is nothing we can really do, and IMO that would still be OK for
most users because squashing has been designed for non-Var lists as
far as I know.  This is a rare pattern for long lists in IN arrays, as
well.  I'll sleep on it for now.
--
Michael

--Pygv+12Q3n0rK3+G
Content-Type: application/pgp-signature; name="signature.asc"

-----BEGIN PGP SIGNATURE-----
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=0KnE
-----END PGP SIGNATURE-----

--Pygv+12Q3n0rK3+G--