From: Tom Lane <tgl@sss.pgh.pa.us>
To: Michael Paquier <michael@paquier.xyz>
cc: exclusion@gmail.com, pgsql-bugs@lists.postgresql.org
Subject: Re: BUG #18135: Incorrect memory access occurs when attaching a
 partition with an index
In-reply-to: <ZRi96-kUPFy0ILzb@paquier.xyz>
References: <18135-532f4a755e71e4d2@postgresql.org>
 <1043711.1695922299@sss.pgh.pa.us> <ZRi96-kUPFy0ILzb@paquier.xyz>
Comments: In-reply-to Michael Paquier <michael@paquier.xyz>
	message dated "Sun, 01 Oct 2023 09:31:39 +0900"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <1614548.1696120777.1@sss.pgh.pa.us>
Date: Sat, 30 Sep 2023 20:39:37 -0400
Message-ID: <1614549.1696120777@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/1614549.1696120777%40sss.pgh.pa.us>
Precedence: bulk

Michael Paquier <michael@paquier.xyz> writes:
> I can see that this has already been committed as 9f71e10d65, but are
> you sure that this is correct?  I didn't take the time to reply back,
> because I got the feeling that even what I proposed is not correct.
> The previous code was careful enough to look at the information from
> info2 only *through* the attribute map, which is not what this new
> code is doing.  Rather than looking directly at the elements in info2
> at each iteration, shouldn't this stuff loop over the elements in
> attmap->attnums for each info1->ii_IndexAttrNumbers[i]?

I think it's OK as is.  What we are comparing in the modified logic
is whether index columns at the same column positions are expressions or
not, and that's not a matter for mapping.  Once we've verified that
the current column is a non-expression in both indexes, then it's
appropriate to use the attmap to see whether the columns correspond.

(Or to put it another way: running "column zero" through the
attmap is always the wrong thing.)

			regards, tom lane