Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm
List-Subscribe: <mailto:cygwin-subscribe AT cygwin DOT com>
List-Archive: <http://sources.redhat.com/ml/cygwin/>
List-Post: <mailto:cygwin AT cygwin DOT com>
List-Help: <mailto:cygwin-help AT cygwin DOT com>, <http://sources.redhat.com/ml/#faqs>
Sender: cygwin-owner AT cygwin DOT com
Mail-Followup-To: cygwin AT cygwin DOT com
Delivered-To: mailing list cygwin AT cygwin DOT com
Message-ID: <3FCB9DDB.4000202@m4x.org>
Date: Mon, 01 Dec 2003 12:00:27 -0800
From: Antoine Labour <antoine DOT labour AT m4x DOT org>
User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.5) Gecko/20031013 Thunderbird/0.3
X-Accept-Language: en-us, en
MIME-Version: 1.0
To: Nicholas Wourms <nwourms AT myrealbox DOT com>
CC: cygwin AT cygwin DOT com
Subject: Re: Bash wait indefinitely
References: <bq720i$b5b$1 AT sea DOT gmane DOT org> <3FC7AC79 DOT 2010000 AT myrealbox DOT com>
In-Reply-To: <3FC7AC79.2010000@myrealbox.com>
Content-Transfer-Encoding: 7bit
Content-Type: text/plain; charset=us-ascii; format=flowed

Nicholas Wourms wrote:
> news.gmane.org wrote:
> 
>> I'm running in concurrences 5 complex bash batch, and sometimes (2 
>> times on
>> 3) one or more (very rarely) batch stop to do something. I've put a 
>> lot of
>> trace to see where is the problem, but seems rarely arrived at the 
>> same line
>> of code.
>>
>> I've also tried to use strace tool, but strace hang rapidly.
>>
> 
> For the last few weeks I've been experiencing the same, and the 
> offenders always seem to be either bash (not ash) or make.  It is 
> important to note that I track cygwin's CVS head, so I had chalked it up 
> to the on-going signal work.  I've found that running deep `make`s or 
> running the autotools in succession (such as in the case of triggering 
> the AM_MAINTAINER_MODE routines) often triggers this problem for me on 
> my Dual Athlon MP 2400+ Win2K box :-(.  I've been extremely busy working 
> on other things, I just grumble, kill the deadlocked process and its 
> children, then restart the parent.  Attempting to debug gdb/strace, as 
> you've discovered, is quite fruitless.
> 
> Cheers,
> Nicholas

Do you guys have things like command substitution in your scripts ? I 
talked about a similar problem in a previous mail (but got no answer), 
it certainly involves some kind of race condition with forks and pipes. 
Going back to a previous version of cygwin (1.3.something) fixes that 
particular problem, but exposes a lot of other pthread races. For my 
part I've been able to attach gdb to the dead processes, but apart from 
dumping the stack(s) traces, I couldn't tell much.

Antoine


--
Unsubscribe info:      http://cygwin.com/ml/#unsubscribe-simple
Problem reports:       http://cygwin.com/problems.html
Documentation:         http://cygwin.com/docs.html
FAQ:                   http://cygwin.com/faq/