Re: chkpass_in should not be volatile - Mailing list pgsql-hackers

From David G. Johnston
Subject Re: chkpass_in should not be volatile
Date
Msg-id CAKFQuwbc56pU1SHpEEhxVhn6Cbs-W-vyRZ1GAmkvNnRF30M5bg@mail.gmail.com
Whole thread Raw
In response to Re: chkpass_in should not be volatile  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: chkpass_in should not be volatile  (Thom Brown <thom@linux.com>)
List pgsql-hackers
On Fri, Jun 3, 2016 at 9:56 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
Thom Brown <thom@linux.com> writes:
> ...or at least according to the warning message:
> postgres=# CREATE EXTENSION chkpass ;
> WARNING:  type input function chkpass_in should not be volatile

See thread here:

https://www.postgresql.org/message-id/flat/CACfv%2BpL2oX08SSZSoaHpyC%3DUbfTFmPt4UmVEKJTH7y%3D2QMRCBw%40mail.gmail.com

Given the lack of complaints so far, maybe we could think about redefining
the behavior of chkpass_in.  I'm not very sure to what, though.

Thom, how did you end up encountering this?

​While it seems to have resulted in the right effect (here) maybe we could have written: "WARNING: If you are reading this please email pgsql-bugs@postgresql.org" and mention checkpass_in volatility in the subject.​" instead

David J.

pgsql-hackers by date:

Previous
From: Tom Lane
Date:
Subject: Re: [BUGS] BUG #14155: bloom index error with unlogged table
Next
From: Noah Misch
Date:
Subject: Re: pg9.6 segfault using simple query (related to use fk for join estimates)