Mailing-List:	contact cygwin-developers-help AT cygwin DOT com; run by ezmlm
List-Subscribe:	<mailto:cygwin-developers-subscribe AT cygwin DOT com>
List-Archive:	<http://sources.redhat.com/ml/cygwin-developers/>
List-Post:	<mailto:cygwin-developers AT cygwin DOT com>
List-Help:	<mailto:cygwin-developers-help AT cygwin DOT com>, <http://sources.redhat.com/ml/#faqs>
Sender:	cygwin-developers-owner AT cygwin DOT com
Delivered-To:	mailing list cygwin-developers AT cygwin DOT com
Message-ID:	<010001c2233f$75231360$6132bc3e@BABEL>
From:	"Conrad Scott" <Conrad DOT Scott AT dsl DOT pipex DOT com>
To:	"Robert Collins" <robert DOT collins AT syncretize DOT net>
Cc:	<cygwin-developers AT cygwin DOT com>
References:	<20020704052424 DOT GA15450 AT redhat DOT com> <00e601c2232d$a7c5be10$2300a8c0 AT LAPTOP>
Subject:	Re: access port 127.0.0.1:1052 (cygserver question) <PLEASE REPLY TO THIS ONE>
Date:	Thu, 4 Jul 2002 10:44:54 +0100
MIME-Version:	1.0
X-Priority:	3
X-MSMail-Priority:	Normal
X-MimeOLE:	Produced By Microsoft MimeOLE V6.00.2600.0000

"Robert Collins" <robert DOT collins AT syncretize DOT net> wrote:
> Sounds like the cygserver domain socket was left open
> (perhaps due to cygserver being forcibly killed),
> and each process was trying to connect to the cygserver.

Rob,

AFAICS it's simply the presence of the socket file (/tmp/cygdaemo)
that triggers this behaviour, i.e. it doesn't require anything to go
wrong to trigger this.  You can re-start the machine and the presence
of the file causes clients to attempt to connect to the daemon.

In Un*x, are socket files persistent in this way? Or do they disappear
on last close?  I can't see anything unambiguous about this in the
documentation I've got to hand.  (I've never programmed with them in
anger and I've no Un*x box to hand to check.)  I could go read the
NetBSD or Linux source some more, but I've got other things I'd like
to do this week :-)

As I mentioned in my reply to Chris, cygserver could unlink the socket
file on exit: this isn't what (for example) XEmacs does with its unix
domain sockets, but I don't know of any good reason why cygserver
*shouldn't* do this.

Any thoughts?

// Conrad

- Raw text -