Обсуждение: Re: [PATCHES] Win32 CHECK_FOR_INTERRUPTS() performance

> > This patch passes regression and demonstrates the expected
> speedup on
> > my machine.
> >
>
> Looks good from here:
>     - several rounds of regression test
>     - psql -f set_time_out.sql
>     - pg_ctl signal sending test
>     - psql deadlock test
>
> Is there any way to test socket?

Send the backend a signal when it's blocking for socket input (waiting
for a user command in a psql session for example), and see that it's
delivered. Hitting hte postmaster will test the select() path, and
hitting a backend will test the recv() path, IIRC.

//Magnus

On Mon, 24 Oct 2005, Magnus Hagander wrote:
> > Is there any way to test socket?
>
> Send the backend a signal when it's blocking for socket input (waiting
> for a user command in a psql session for example), and see that it's
> delivered. Hitting hte postmaster will test the select() path, and
> hitting a backend will test the recv() path, IIRC.
>

Ok.
- Test recv() path by runing a psql with "-f time_out.sql" and a batch
file executing "pg_ctl kill HUP/USR1" repeatedly.- Test select() path by running two batch files, one is doing "psql
-c
"set statement_timeout=10" test", the other is doing "pg_ctl kill HUP
<postmaster>".

Regards,
Qingqing