On Thu, Jun 30, 2016 at 9:35 PM, Fujii Masao <masao.fujii@gmail.com> wrote:
> On Thu, Jun 30, 2016 at 9:30 PM, Michael Paquier
> <michael.paquier@gmail.com> wrote:
>> On Thu, Jun 30, 2016 at 8:59 PM, Fujii Masao <masao.fujii@gmail.com> wrote:
>>> (2)
>>> +retry:
>>> + SpinLockAcquire(&walrcv->mutex);
>>> + if (!walrcv->ready_to_display)
>>> + {
>>> + SpinLockRelease(&walrcv->mutex);
>>> + CHECK_FOR_INTERRUPTS();
>>> + pg_usleep(1000);
>>> + goto retry;
>>> + }
>>> + SpinLockRelease(&walrcv->mutex);
>>>
>>> ISTM that we will never be able to get out of this loop if walreceiver
>>> fails to connect to the master (e.g., password is wrong) after we enter
>>> this loop.
>>
>> Wouldn't it be cleaner to just return an error here instead of retrying?
>
> I prefer to return NULL. Now NULL is returned when walreceiver's pid is 0.
> We can just change this logic so that NULL is returned pid is 0 OR the
> flag is false.
OK, yes. That's indeed better this way. Need a patch?
--
Michael
