Re: Minor comment update in execPartition.c

Поиск
Список
Период
Сортировка
От Amit Langote
Тема Re: Minor comment update in execPartition.c
Дата
Msg-id 78e033fb-56ef-38e7-2ca3-19e8fff77990@lab.ntt.co.jp
обсуждение исходный текст
Ответ на Minor comment update in execPartition.c  (Etsuro Fujita <fujita.etsuro@lab.ntt.co.jp>)
Ответы Re: Minor comment update in execPartition.c  (Alvaro Herrera <alvherre@alvh.no-ip.org>)
Список pgsql-hackers
Fujita-san,

On 2018/04/24 20:14, Etsuro Fujita wrote:
> Hi,
> 
> Here is a comment for ExecInitPartitionInfo:
> 
>  296  * ExecInitPartitionInfo
>  297  *      Initialize ResultRelInfo and other information for a
> partition if not
>  298  *      already done
> 
> I think we should remove the words "if not already done" from that
> comment because 1) that function is called if the partition wasn't
> already initialized and 2) that function assumes that.  Attached is a
> small patch for removing the words.

Thanks, sounds fine.  I think those words remain from earlier versions of
the patch committed as edd44738bc8 [1], where there used to be a fast-path
return if the ResultRelInfo was already non-NULL for the passed in index.
 I remember you suggested making it so that we call ExecInitPartitionInfo
only if needed [2].  After making the changes as suggested, the "if not
already done" became redundant.

Thanks,
Amit

[1] https://git.postgresql.org/gitweb/?p=postgresql.git;a=commit;h=edd44738bc8

[2] https://www.postgresql.org/message-id/5A7443DF.1010408%40lab.ntt.co.jp



В списке pgsql-hackers по дате отправления:

Предыдущее
От: Peter Eisentraut
Дата:
Сообщение: Re: obsoleting plpython2u and defaulting plpythonu to plpython3u
Следующее
От: Kyotaro HORIGUCHI
Дата:
Сообщение: Re: Oddity in tuple routing for foreign partitions