Mailing-List: contact cygwin-help@sourceware.cygnus.com; run by ezmlm
List-Subscribe: <mailto:cygwin-subscribe@sources.redhat.com>
List-Archive: <http://sources.redhat.com/ml/cygwin/>
List-Post: <mailto:cygwin@sources.redhat.com>
List-Help: <mailto:cygwin-help@sources.redhat.com>, <http://sources.redhat.com/ml/#faqs>
Sender: cygwin-owner@sources.redhat.com
Delivered-To: mailing list cygwin@sources.redhat.com
Message-ID: <F10D23B02E54D011A0AB0020AF9CEFE988FA5E@lynx.ceddec.com>
From: "Town, Brad" <btown@ceddec.com>
To: "'Joost Kraaijeveld'" <J.Kraaijeveld@Askesis.nl>,
        "'cygwin@cygwin.com'"
	 <cygwin@cygwin.com>
Subject: RE: Exception problems-> this is a bug!
Date: Tue, 28 Nov 2000 11:08:29 -0500
MIME-Version: 1.0
X-Mailer: Internet Mail Service (5.5.2653.19)
Content-Type: text/plain;
	charset="iso-8859-1"

I compiled using "g++ -g test.cc", and I'm seeing the behavior that Joost
mentioned.  gdb caught the exception and is currently pointing at an
assembly instruction between the last brace of the for loop and the last
brace of the try block: "mov (%edx),%ecx".

I did have to tweak the source a bit to get it to run.  On my machine, I got
the message "couldn't dynamically determine load address for
'CharToOemBuffA' (handle 0x0), Win32 error 127".  Putting a Sleep between
the two calls to CreateThread allows the program to run.  Looks like the
assembly in autoload.cc needs to be thread-safe (just like the FIXME above
it says).

Brad Town


Joost Kraaijeveld [mailto:J.Kraaijeveld@Askesis.nl] wrote:
> This program (see below) crashes whenever a function on a 
> separate thread is
> called that contains a try{} catch(){} block. I tested with several
> compilers (Cygwin, MingW, IBM Visual Age, MS Visual C++ and 
> Metrowerks) and
> only CygWin (latest release of all software) crashes. I think 
> that the stack
> of the thread is not cleaned up correctly.
> 
> Is there a workaround availalbe of some kind??
<snip>

--
Want to unsubscribe from this list?
Send a message to cygwin-unsubscribe@sourceware.cygnus.com