Mailing-List: contact cygwin-help@sourceware.cygnus.com; run by ezmlm
Sender: cygwin-owner@sourceware.cygnus.com
From: =?iso-8859-1?Q?Boris_Sch=E4ling?= <boriss@web.de>
To: <cygwin@sourceware.cygnus.com>
Subject: RE: select() implementation question
Date: Tue, 20 Jun 2000 16:31:41 +0200
Message-ID: <LOBBKDBPKEGJJFKNFHKBIECKDEAA.boriss@web.de>
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: 7bit
Importance: Normal
In-Reply-To: <20000619201147.A936@cygnus.com>


> -----Original Message-----
> From: cygwin-owner@sourceware.cygnus.com
> [mailto:cygwin-owner@sourceware.cygnus.com]On Behalf Of Chris Faylor
> Sent: Tuesday, June 20, 2000 2:12 AM
> To: cygwin@sourceware.cygnus.com
> Subject: Re: select() implementation question
>
> [...]
> Taking a quick glance at select.cc, I can see that cygwin select does
> open a dummy socket.  Possibly it is not closing something when it
> returns.  I really don't know.
Ah, a dummy socket. Well, then I think this is not a bug but a feature: I
guess this socket is used to jump back from select() whenever cygwin needs
to do so. There are probably some things to do that can't wait until a
socket is ready or the timer decreased to 0. I think this additional socket
is used to have some sort of control and to interrupt select() in a proper
way.

However, I downloaded the source code of select(). I'll have a look when I
have some time.

BTW: Is anybody working already on an implementation of getaddrinfo(),
gai_strerror() and freeaddrinfo()? Otherwise I will try. :-)

Boris


--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com