From: Tom Lane <tgl@sss.pgh.pa.us>
To: suresh.arsenal29@gmail.com
cc: pgsql-bugs@lists.postgresql.org
Subject: Re: BUG #15795: ERROR: could not find pathkey item to sort
In-reply-to: <15795-fadb56c8e44ee73c@postgresql.org>
References: <15795-fadb56c8e44ee73c@postgresql.org>
Comments: In-reply-to PG Bug reporting form <noreply@postgresql.org>
	message dated "Wed, 08 May 2019 06:51:32 -0000"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <18170.1557311598.1@sss.pgh.pa.us>
Content-Transfer-Encoding: quoted-printable
Date: Wed, 08 May 2019 06:33:18 -0400
Message-ID: <18171.1557311598@sss.pgh.pa.us>
Precedence: bulk

PG Bug reporting form <noreply@postgresql.org> writes:
> Here is query:
> SELECT  DISTINCT  A._id0 as _id0, A._actual_type_name0 as _actual_type_n=
ame0
>  FROM  ( (   SELECT  DISTINCT _id as _id0, _actual_type_name as
> _actual_type_name0, name as name0  FROM  hello_world.person  ) union all=
  ( =

> SELECT  DISTINCT  _id as _id0, _actual_type_name as _actual_type_name0, =
name
> as name0  FROM  hello_world.person)) as A WHERE ( A.name0  =3D  A.name0 =
);

It's politer to provide a self-contained test case, rather than expect us
to reverse-engineer details that might be critical.

For the archives, though, this isn't hard to reproduce:

regression=3D# create table person(_id int, _actual_type_name text, name t=
ext);
CREATE TABLE
regression=3D# SELECT  DISTINCT  A._id0 as _id0, A._actual_type_name0 as _=
actual_type_name0
 FROM  ( (   SELECT  DISTINCT _id as _id0, _actual_type_name as
_actual_type_name0, name as name0  FROM  person  ) union all  ( =

SELECT  DISTINCT  _id as _id0, _actual_type_name as _actual_type_name0, na=
me
as name0  FROM  person)) as A WHERE ( A.name0  =3D  A.name0 );
ERROR:  could not find pathkey item to sort

Curiously, this only fails for me in 9.6 and 10, not earlier or later
branches.

			regards, tom lane