Horton

Re: Wrong aggregate result when sorting by a NULL value

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: Jeff Janes <jeff.janes@gmail.com>, Andrew Gierth <andrew@tao11.riddles.org.uk>, obouda@email.cz, Pg Bugs <pgsql-bugs@postgresql.org>
Date: 2018-11-05T00:15:47Z
Lists: pgsql-bugs 
Andres Freund <andres@anarazel.de> writes:
> On 2018-11-04 18:49:40 -0500, Tom Lane wrote:
>> Meh.  I'm unexcited about getting rid of one type of compiler warning by
>> introducing another one.

> It's impracticable to enable dead code warnings in postgres
> anyway. There's way way too many of them (all the returns after
> elog(ERROR) etc are enough to make it infeasible to change that).

I think you have a very limited conception of what sorts of warnings
code like that would trigger.  In particular, even given the
unsubstantiated assumption that nobody's compiler would complain,
I think static analyzers like Coverity might.

			regards, tom lane

Commits

  1. Fix unused-variable warning.

  2. Prevent generating EEOP_AGG_STRICT_INPUT_CHECK operations when nargs == 0.

  3. Fix STRICT check for strict aggregates with NULL ORDER BY columns.

  4. Expression evaluation based aggregate transition invocation.