Mailing-List: contact cygwin-help AT sourceware DOT cygnus DOT com; run by ezmlm
List-Subscribe: <mailto:cygwin-subscribe AT sources DOT redhat DOT com>
List-Archive: <http://sources.redhat.com/ml/cygwin/>
List-Post: <mailto:cygwin AT sources DOT redhat DOT com>
List-Help: <mailto:cygwin-help AT sources DOT redhat DOT com>, <http://sources.redhat.com/ml/#faqs>
Sender: cygwin-owner AT sources DOT redhat DOT com
Delivered-To: mailing list cygwin AT sources DOT redhat DOT com
Date: Wed, 27 Jun 2001 21:22:01 -0400
From: Christopher Faylor <cgf AT redhat DOT com>
To: cygwin AT cygwin DOT com
Subject: Re: Deadly embrace between pthread_cond_wait and pthread_cond_signal
Message-ID: <20010627212201.A27163@redhat.com>
Reply-To: cygwin AT cygwin DOT com
Mail-Followup-To: cygwin AT cygwin DOT com
References: <20010627023524 DOT S19058 AT redhat DOT com> <00aa01c0ff1b$44106960$a300a8c0 AT nhv> <20010627112321 DOT A21615 AT redhat DOT com> <037401c0ff5c$ed60ebc0$806410ac AT local> <3B3A86CF DOT DC0FA39A AT nc DOT rr DOT com> <3B3A87BB DOT 3F7AF46 AT nc DOT rr DOT com>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
User-Agent: Mutt/1.3.11i
In-Reply-To: <3B3A87BB.3F7AF46@nc.rr.com>; from gsmith@nc.rr.com on Wed, Jun 27, 2001 at 09:26:19PM -0400

On Wed, Jun 27, 2001 at 09:26:19PM -0400, Greg Smith wrote:
>Robert Collins wrote:
>
>>Volume of calls :}.  openening files doesn't occur quite as often as
>>(say)
>^^^^^^^^^ That might be easy for you to say !!
>
>>I'd like to publicly thank Greg Smith for his excellent profiling work
>>in identifying the bottleneck here.  I know how time consuming that can
>>be :}.
>
>That's all fine and dandy, what I wanna know is `When's the check
>coming in the mail ?' ;-)
>
>I do have another problem, which I think is associated with
>pthread_kill, signals, and related ilk.
>
>Seems we do a lotta SIGHUPs to break a listening thread out of his
>reverie, but sometimes the wrong thead (which has done a recv() on a
>socket) gets woken up instead, and is not particularly happy with the
>EINTR error code.

Unless Robert has performed some magic that I'm not aware of, there is
currently no way to interrupt anything but the main thread in cygwin.

This has been on my tuit list for a long time but it will require a
lot of rework of the signal code.

cgf

--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Bug reporting:         http://cygwin.com/bugs.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/