Re: Remove some redundant set_cheapest() calls

Поиск
Список
Период
Сортировка
От Richard Guo
Тема Re: Remove some redundant set_cheapest() calls
Дата
Msg-id CAMbWs4_zhKyQaxOFe8mViY_fsi9rO=wsALijRg+PvXvsr9=U+w@mail.gmail.com
обсуждение исходный текст
Ответ на Re: Remove some redundant set_cheapest() calls  (Tom Lane <tgl@sss.pgh.pa.us>)
Ответы Re: Remove some redundant set_cheapest() calls  (Tom Lane <tgl@sss.pgh.pa.us>)
Список pgsql-hackers

On Wed, Mar 27, 2024 at 4:06 AM Tom Lane <tgl@sss.pgh.pa.us> wrote:
Richard Guo <guofenglinux@gmail.com> writes:
> I happened to notice that the set_cheapest() calls in functions
> set_namedtuplestore_pathlist() and set_result_pathlist() are redundant,
> as we've centralized the set_cheapest() calls in set_rel_pathlist().

> Attached is a trivial patch to remove these calls.

Agreed, and pushed.

Thanks for pushing!
 
> BTW, I suspect that the set_cheapest() call in set_dummy_rel_pathlist()
> is also redundant.  The comment there says "This is redundant when we're
> called from set_rel_size(), but not when called from elsewhere".  I
> doubt it.  The other places where it is called are set_append_rel_size()
> and set_subquery_pathlist(), both being called from set_rel_size().  So
> set_cheapest() would ultimately be called in set_rel_pathlist().

I'm less convinced about changing this.  I'd rather keep it consistent
with mark_dummy_rel.

Hm, I wonder if we should revise the comment there that states "but not
when called from elsewhere", as it does not seem to be true.

Thanks
Richard

В списке pgsql-hackers по дате отправления:

Предыдущее
От: Donghang Lin
Дата:
Сообщение: Re: Parallel Bitmap Heap Scan reports per-worker stats in EXPLAIN ANALYZE
Следующее
От: Bertrand Drouvot
Дата:
Сообщение: Re: pgsql: Track last_inactive_time in pg_replication_slots.