Re: New win32 signals patch (3)

From: "Magnus Hagander" <mha(at)sollentuna(dot)net>
To: "Tom Lane" <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Cc: "Claudio Natoli" <claudio(dot)natoli(at)memetrics(dot)com>, "pgsql-hackers-win32" <pgsql-hackers-win32(at)postgresql(dot)org>
Subject: Re: New win32 signals patch (3)
Date: 2004-02-04 16:28:01
Message-ID: 6BCB9D8A16AC4241919521715F4D8BCE34B12C@algol.sollentuna.se
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers-win32

> > The limit for select on win32 is actually 64, which is
> horribly low in
> > many cases.
>
> Wow, that small and they don't have poll()? Whatta bunch of bozos.

Well, they have other methods. Such as their "I/O Completion Ports",
which I've seen easily handle 30-40,000 sockets (IIRC that's the number
- it was at least very large)... But they do like to do things their own
way...

> > But as it is right now, select() is only used in the
> postmaster and in
> > the pgstat process, neither of which use any large number of fds.
> > Since we only select() on sockets, and not the files.
>
> Okay. I thought I had seen something about using select() in
> the context of catching signals, in which case it'd be
> necessary to use it in all backends. But if not then we can
> probably live with the small fd_set size.

If you're talking win32 port specific, I think it was the other way
around. Using select() *prevents* catching signals with the current
implementation. This is why we need the "select replacement", see my
latest patch.

//Magnus

Browse pgsql-hackers-win32 by date

  From Date Subject
Next Message Bruce Momjian 2004-02-04 18:41:03 Re: Sync vs. fsync during checkpoint
Previous Message Tom Lane 2004-02-04 16:18:20 Re: New win32 signals patch (3)