From: Antonin Houska <ah@cybertec.at>
To: Chao Li <li.evan.chao@gmail.com>
cc: PostgreSQL-development <pgsql-hackers@postgresql.org>,
    "Zhijie Hou (Fujitsu)" <houzj.fnst@fujitsu.com>,
    Alvaro Herrera <alvherre@alvh.no-ip.org>
Subject: Re: repack: fix a bug to reject deferrable primary key fallback for
 concurrent mode
In-reply-to: <48844EC6-0909-41E6-B2C4-F3BF8A2B1AA8@gmail.com>
References: <10DD5E13-B45D-44F1-BE08-C63E00ABCAC0@gmail.com>
 <65564.1776696735@localhost> <48844EC6-0909-41E6-B2C4-F3BF8A2B1AA8@gmail.com>
Comments: In-reply-to Chao Li <li.evan.chao@gmail.com>
   message dated "Tue, 21 Apr 2026 10:38:05 +0800."
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <5989.1776751782.1@localhost>
Content-Transfer-Encoding: quoted-printable
Date: Tue, 21 Apr 2026 08:09:42 +0200
Message-ID: <5990.1776751782@localhost>
Archived-At: 
 <https://www.postgresql.org/message-id/5990.1776751782%40localhost>
Precedence: bulk

Chao Li <li.evan.chao@gmail.com> wrote:

> > On Apr 20, 2026, at 22:52, Antonin Houska <ah@cybertec.at> wrote:
> > =

> > I'm just thinking if it's worth a separate error message.
> > RelationGetIndexList() just ignores the deferrable PK
> > =

> > if (replident =3D=3D REPLICA_IDENTITY_DEFAULT && OidIsValid(pkeyIndex)=
 && !pkdeferrable)
> > relation->rd_replidindex =3D pkeyIndex;
> > =

> > and if there's no other suitable index, the result is that there is no
> > identity index for the table. So the change attached here should be co=
nsistent
> > with this approach.

> Thanks for your review. I guess you read the v1 patch. In v2, I have swi=
tched to use GetRelationIdentityOrPK() that Zhijie suggested, which has co=
vered RelationGetIndexList() and all checks, so that code is simplified, a=
nd there is no longer a separate error message.

Yes, this looks like the best approach. Sorry for missing v2.

-- =

Antonin Houska
Web: https://www.cybertec-postgresql.com