Horton

Re: strange failure in plpgsql_control tests (on fulmar, ICC 14.0.3)

Tom Lane <tgl@sss.pgh.pa.us>

View thread View on postgr.es
From: Tom Lane <tgl@sss.pgh.pa.us>
To: Andres Freund <andres@anarazel.de>
Cc: pgsql-hackers@lists.postgresql.org, Tomas Vondra <tomas.vondra@2ndquadrant.com>, "pgsql-hackers@postgresql.org" <pgsql-hackers@postgresql.org>
Date: 2018-03-17T18:32:36Z
Lists: pgsql-hackers 
Andres Freund <andres@anarazel.de> writes:
> On the current branch just using the new overflow safe functions in
> int.h should work. But unless we are OK leaving this broken in the back
> branches, or want to backport the functionality, that's probably not
> sufficient.

Yeah ... I don't like either of the last two things, so probably we should
go with the patch as I had it.  Yours might perform a shade better on
compilers with the built-in, but it'll be a lot worse on those without.

What I was wondering about was whether to back-patch a test case.
It doesn't seem really necessary, and we'd have to put it someplace
else than where it is in HEAD, so I'm leaning against.

			regards, tom lane

Commits

  1. Fix overflow handling in plpgsql's integer FOR loops.