Re: GROUP BY DISTINCT

From: Tomas Vondra <tomas(dot)vondra(at)enterprisedb(dot)com>
To: Thomas Munro <thomas(dot)munro(at)gmail(dot)com>
Cc: Vik Fearing <vik(at)postgresfriends(dot)org>, Georgios Kokolatos <gkokolatos(at)protonmail(dot)com>, PostgreSQL Hackers <pgsql-hackers(at)lists(dot)postgresql(dot)org>, Erik Rijkers <er(at)xs4all(dot)nl>
Subject: Re: GROUP BY DISTINCT
Date: 2021-03-18 21:14:03
Message-ID: aa87b9b9-6a14-a84a-574f-e3ad2ce2504c@enterprisedb.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-docs pgsql-hackers

On 3/18/21 10:02 PM, Thomas Munro wrote:
> On Fri, Mar 19, 2021 at 8:27 AM Tomas Vondra
> <tomas(dot)vondra(at)enterprisedb(dot)com> wrote:
>> Hmmm, this seems to fail on lapwing with this error:
>>
>> parse_agg.c: In function 'expand_grouping_sets':
>> parse_agg.c:1851:23: error: value computed is not used
>> [-Werror=unused-value]
>> cc1: all warnings being treated as errors
>>
>> That line is this:
>>
>> foreach_delete_current(result, cell);
>>
>> and I don't see how any of the values close by could be unused ...
>>
>> The only possibility I can think of is some sort of issue in the old-ish
>> gcc release (4.7.2).
>
> No sure what's going on there, but data points: I tried a 32 bit build
> here (that's the other special thing about lapwing) and didn't see the
> warning. GCC 10. Also curculio (gcc 4.2) and snapper (gcc 4.7) are
> also showing this warning, but they don't have -Werror so they don't
> fail. sidewinder (gcc 4.8) is not showing the warning.
>

Thanks for the info. So it's likely related to older gcc releases. The
question is how to tweak the code to get rid of this ...

regards

--
Tomas Vondra
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company

In response to

Responses

Browse pgsql-docs by date

  From Date Subject
Next Message Thomas Munro 2021-03-18 21:57:13 Re: GROUP BY DISTINCT
Previous Message Thomas Munro 2021-03-18 21:02:53 Re: GROUP BY DISTINCT

Browse pgsql-hackers by date

  From Date Subject
Next Message Matthias van de Meent 2021-03-18 21:30:36 Re: non-HOT update not looking at FSM for large tuple update
Previous Message Bruce Momjian 2021-03-18 21:10:12 Re: Maintaining a list of pgindent commits for "git blame" to ignore