X-Spam-Check-By: sourceware.org
Date: Wed, 22 Mar 2006 13:55:41 +0100
From: Corinna Vinschen <corinna-cygwin AT cygwin DOT com>
To: cygwin AT cygwin DOT com
Subject: Re: pthread_create leaves valid mutex pointers on the stack
Message-ID: <20060322125541.GB9091@calimero.vinschen.de>
Reply-To: cygwin AT cygwin DOT com
Mail-Followup-To: cygwin AT cygwin DOT com
References: <17433 DOT 39495 DOT 950903 DOT 287320 AT metro-north DOT cs DOT columbia DOT edu> <20060316171139 DOT GM4280 AT implementation DOT labri DOT fr>
Mime-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <20060316171139.GM4280@implementation.labri.fr>
User-Agent: Mutt/1.4.2i
Mailing-List: contact cygwin-help AT cygwin DOT com; run by ezmlm
Precedence: bulk
List-Unsubscribe: <mailto:cygwin-unsubscribe-archive-cygwin=delorie DOT com AT cygwin DOT com>
List-Subscribe: <mailto:cygwin-subscribe AT cygwin DOT com>
List-Archive: <http://sourceware.org/ml/cygwin/>
List-Post: <mailto:cygwin AT cygwin DOT com>
List-Help: <mailto:cygwin-help AT cygwin DOT com>, <http://sourceware.org/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

On Mar 16 18:11, Samuel Thibault wrote:
> Hi,
> 
> Jonathan Lennox, le Thu 16 Mar 2006 12:03:03 -0500, a écrit :
> Content-Description: message body text
> > In general, the idea of verifying objects on their init functions seems
> > dubious to me -- how can you tell initialized objects from random stack or
> > heap garbage?
> 
> Posix says:
>     Attempting to initialize an already initialized mutex results in undefined
>     behavior.

According to SUSv3 pthread_mutex_init is allowed to check the incoming
mutex and return EBUSY if the mutex is already in use:

  The pthread_mutex_init() function may fail if:

  [EBUSY]
      The implementation has detected an attempt to reinitialize the
      object referenced by mutex, a previously initialized, but not yet
      destroyed, mutex.

But the problem is obvious.  On initializing an object you can't trust
the value on the stack to be meant as a pointer to an existing object,
or to be just random bytes or, in your nice example, being a value left
on the stack by an earlier function call.  That's probably the reason
why SUSv3 says "may" and not "must", and that's probably also the reason
that such a test isn't made on Linux.  I disabled the validity checks
now in the methods corresponding to the functions pthread_mutex_init,
pthread_rwlock_init and pthread_cond_init.

Thanks for the testcase, Jonathan.


Corinna


-- 
Corinna Vinschen                  Please, send mails regarding Cygwin to
Cygwin Project Co-Leader          cygwin AT cygwin DOT com
Red Hat

--
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/