Mailing-List: contact cygwin-help AT sourceware DOT cygnus DOT com; run by ezmlm Sender: cygwin-owner AT sourceware DOT cygnus DOT com Delivered-To: mailing list cygwin AT sourceware DOT cygnus DOT com X-Authentication-Warning: modi.xraylith.wisc.edu: khan owned process doing -bs Date: Wed, 12 May 1999 17:08:49 -0500 (CDT) From: Mumit Khan To: Isselmou Ould-Dellahy cc: cygwin AT sourceware DOT cygnus DOT com Subject: Re: ****** /home/noer/src/b20/comp-tools/devo/winsup/exceptions.cc In-Reply-To: <19990511205631.A28641@eastern.ccrl.mot.com> Message-ID: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On Tue, 11 May 1999, Isselmou Ould-Dellahy wrote: > Hi, > Could anyone give me an explanation for this message : > > ****** /home/noer/src/b20/comp-tools/devo/winsup/exceptions.cc:348 __get_signal_mutex failed, res -1, error 6 > > This is a runtime error sent by a java/C/C++ code before it crashes. > I tried to minimise the amount of code that produces it but could not; > the crash seems arbitrary. Cygwin is not thread safe, and I don't think it's safe currently to use Cygwin for any non-trivial JNI. I've noticed that it seems to work as long as I limit myself to a single JNI, but as soon as I load more than one, kabooom (and seemingly at random places). Can you try with the Mingw runtime and see if the problem persists? > BTW : what is the best debugger for such a software (a java > gui that calls C native methods ) ? GDB can do the job, albeit in a somewhat awkward manner. You essentially run java under gdb and then it'll stop at the crash. Regards, Mumit -- Want to unsubscribe from this list? Send a message to cygwin-unsubscribe AT sourceware DOT cygnus DOT com