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
content-class: urn:content-classes:message
Subject: RE: Deadly embrace between pthread_cond_wait and pthread_cond_signal
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Date: Thu, 28 Jun 2001 12:08:59 +1000
X-MimeOLE: Produced By Microsoft Exchange V6.0.4417.0
Message-ID: <EA18B9FA0FE4194AA2B4CDB91F73C0EF08F0AD@itdomain002.itdomain.net.au>
X-MS-Has-Attach: 
X-MS-TNEF-Correlator: 
Thread-Topic: Deadly embrace between pthread_cond_wait and pthread_cond_signal
Thread-Index: AcD/cAPw8WGjIBmWQ5aHwPHMu6JQhQABwyKg
From: "Robert Collins" <robert DOT collins AT itdomain DOT com DOT au>
To: "Greg Smith" <gsmith AT nc DOT rr DOT com>, <cygwin AT cygwin DOT com>
Content-Transfer-Encoding: 8bit
X-MIME-Autoconverted: from quoted-printable to 8bit by delorie.com id WAA08787

> -----Original Message-----
> From: Greg Smith [mailto:gsmith AT nc DOT rr DOT com]
> Sent: Thursday, June 28, 2001 11:26 AM
> To: cygwin AT cygwin DOT com
> Subject: Re: Deadly embrace between pthread_cond_wait and
> pthread_cond_signal
> 
> 
> 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 ?'  ;-)

Right after I get mine :].
 
> 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.
> 
> Soon as I can provide more details, I'll open a new thread ;-)

Please do. As Chris has already mentioned the signals code isn't
particulary thread capable right now. It's thread safe, but that's a
different thing. I don't know the signals code, but I'm happy to attempt
tweaks on it :}.

Rob

--
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/