On 2020/02/17 18:48, Michael Paquier wrote:
> On Mon, Feb 17, 2020 at 04:30:00PM +0900, Fujii Masao wrote:
>> On 2020/02/14 23:43, Robert Haas wrote:
>>> On Thu, Feb 13, 2020 at 10:47 PM Fujii Masao
>>> <masao.fujii@oss.nttdata.com> wrote:
>>>> Fixed. Thanks for the review!
>>>
>>> I think it would be safer to just report the wait event during
>>> pg_usleep(1000000L) rather than putting those calls around the whole
>>> loop. It does not seem impossible that ereport() or
>>> CHECK_FOR_INTERRUPTS() could do something that reports a wait event
>>> internally.
>
> CHECK_FOR_INTERRUPTS() would reset the event wait state. Hm.. You
> may be right about the WARNING and it would be better to not rely on
> that. Do you remember the states which may be triggered?
>
>> OK, so I attached the updated version of the patch.
>> Thanks for the review!
>
> Actually, I have some questions:
> 1) Should a new wait event be added in recoveryPausesHere()? That
> would be IMO useful.
Yes, it's useful, I think. But it's better to implement that
as a separate patch.
> 2) Perhaps those two points should be replaced with WaitLatch(), where
> we would use the new wait events introduced?
For what? Maybe it should, but I'm not sure it's worth the trouble.
Regards,
--
Fujii Masao
NTT DATA CORPORATION
Advanced Platform Technology Group
Research and Development Headquarters