From: Tom Lane <tgl@sss.pgh.pa.us>
To: David Rowley <dgrowleyml@gmail.com>
cc: Dmytro Astapov <dastapov@gmail.com>, pgsql-bugs@lists.postgresql.org
Subject: Re: Surprising SeqScan of appendRel that can't contribute any rows to
 the result
In-reply-to: 
 <CAApHDvp5Rj9KwbyOMBWHghJjmX9=cZkhFHSd8U1e5BLP-LeYEQ@mail.gmail.com>
References: 
 <CAFQUnFh662SNr8pmJt2BDimTwXvMnv-cRBAECAHP5sG7U+6rLw@mail.gmail.com>
 <CAApHDvp5Rj9KwbyOMBWHghJjmX9=cZkhFHSd8U1e5BLP-LeYEQ@mail.gmail.com>
Comments: In-reply-to David Rowley <dgrowleyml@gmail.com>
	message dated "Wed, 18 Dec 2024 13:08:10 +1300"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <2549067.1734486021.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Tue, 17 Dec 2024 20:40:21 -0500
Message-ID: <2549068.1734486021@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/2549068.1734486021%40sss.pgh.pa.us>
Precedence: bulk

David Rowley <dgrowleyml@gmail.com> writes:
> On Wed, 18 Dec 2024 at 12:17, Dmytro Astapov <dastapov@gmail.com> wrote:
>> Surprisingly, this does SeqScan on partB in NestedLoops over some_ids w=
ith a filter `some_ids.id =3D NULL::bigint`:

> TL;DR is it's not a bug and expected behaviour.

The particular case shown here might be fixable by re-applying
eval_const_expressions after we've derived the pushed-down qual.
Most of the time that'd be a waste of cycles though, and I'm
not sure how often it would help.

			regards, tom lane