From xemacs-m  Mon Mar 10 05:51:42 1997
Received: from venus.Sun.COM (venus.Sun.COM [192.9.25.5])
	by xemacs.org (8.8.5/8.8.5) with SMTP id FAA22051
	for <xemacs-beta@xemacs.org>; Mon, 10 Mar 1997 05:51:42 -0600 (CST)
Received: from Eng.Sun.COM ([129.146.1.25]) by venus.Sun.COM (SMI-8.6/mail.byaddr) with SMTP id DAA13311; Mon, 10 Mar 1997 03:51:13 -0800
Received: from kindra.eng.sun.com by Eng.Sun.COM (SMI-8.6/SMI-5.3)
	id DAA03072; Mon, 10 Mar 1997 03:51:11 -0800
Received: from xemacs.eng.sun.com by kindra.eng.sun.com (SMI-8.6/SMI-SVR4)
	id DAA27809; Mon, 10 Mar 1997 03:51:10 -0800
Received: by xemacs.eng.sun.com (SMI-8.6/SMI-SVR4)
	id DAA04976; Mon, 10 Mar 1997 03:51:08 -0800
Date: Mon, 10 Mar 1997 03:51:08 -0800
Message-Id: <199703101151.DAA04976@xemacs.eng.sun.com>
From: Martin Buchholz <mrb@Eng.Sun.COM>
To: David Moore <dmoore@ucsd.edu>
Cc: xemacs-beta@xemacs.org
Subject: Re: crash
In-Reply-To: <rvhgikc28d.fsf@sdnp5.ucsd.edu>
References: <199703091346.HAA16099@mharnois.workgroup.net>
	<rvhgikc28d.fsf@sdnp5.ucsd.edu>
Reply-To: Martin Buchholz <mrb@Eng.Sun.COM>
Mime-Version: 1.0 (generated by tm-edit 7.101)
Content-Type: text/plain; charset=US-ASCII

>>>>> "David" == David Moore <dmoore@ucsd.edu> writes:

David> Michael Harnois <mharnois@sbt.net> writes:
>> This happened when I was trying to save a file.

>> #13 0x808eaf0 in fatal_error_signal (sig=11) at emacs.c:179
>> #14 0xbfffe8d4 in ?? ()
>> #15 0x8055e42 in gc_sweep () at alloc.c:3860
>> #16 0x8056481 in garbage_collect_1 () at alloc.c:4131
>> #17 0x8094f5d in funcall_recording_as (recorded_as=405987108, nargs=3, 
>> args=0xbfffeb04) at eval.c:3093

David> 	Looks like 20.1 from the line numbers for main in emacs.c.
David> Please do say the version, since there are some differences even in the
David> non-MULE parts of the code.

David> 	Looks like there are some serious GC problems.  Is anyone
David> familar with this code?  This looks different than the previously
David> reported problems with Lstream_close being called during GC, since it
David> appears that the function in frame #14 is likely sweep_symbols.

Ben's opinion was that crashes in garbage collection are generally
signs of earlier memory corruption.  The crashes are unreproducible
and the stack traces are unfortunately close to useless.  The best
tool to find these sorts of bugs are the Purify class of memory
debugging tools.

Martin

