Re: Remove extraneous break condition in logical slot advance function

Поиск
Список
Период
Сортировка
От Michael Paquier
Тема Re: Remove extraneous break condition in logical slot advance function
Дата
Msg-id ZTXLTpkSpQ-r4G8J@paquier.xyz
обсуждение исходный текст
Ответ на Re: Remove extraneous break condition in logical slot advance function  (Bharath Rupireddy <bharath.rupireddyforpostgres@gmail.com>)
Список pgsql-hackers
On Sun, Oct 22, 2023 at 11:59:00PM +0530, Bharath Rupireddy wrote:
> AFAICS, there's no correctness argument for breaking before CFI. As
> rightly said, CFIs can happen before the break condition either down
> inside LogicalDecodingProcessRecord or XLogReadRecord (page_read
> callbacks for instance).
>
> Having said that, what may happen if CFI happens and interrupts are
> processed before the break condition is that the decoding occurs again
> which IMV is not a big problem.
>
> An idea to keep all of  XLogReadRecord() -
> LogicalDecodingProcessRecord() loops consistent is by having CFI at
> the start of the loops before the XLogReadRecord().

Passing by..  All that just looks like an oversight of 38a957316d7e
that simplified the main while loop, so I've just applied your v2.
--
Michael

Вложения

В списке pgsql-hackers по дате отправления:

Предыдущее
От: Michael Paquier
Дата:
Сообщение: Re: Show version of OpenSSL in ./configure output
Следующее
От: sirisha chamarthi
Дата:
Сообщение: Re: Why is hot_standby_feedback off by default?