From owner-freebsd-hackers  Thu Oct 12 07:27:31 1995
Return-Path: owner-hackers
Received: (from root@localhost)
          by freefall.freebsd.org (8.6.12/8.6.6) id HAA17227
          for hackers-outgoing; Thu, 12 Oct 1995 07:27:31 -0700
Received: from expo.x.org (expo.x.org [198.112.45.11])
          by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id HAA17206
          for <hackers@freefall.FreeBSD.org>; Thu, 12 Oct 1995 07:27:22 -0700
Received: from exalt.x.org by expo.x.org id AA11876; Thu, 12 Oct 95 10:26:49 -0400
Received: from localhost by exalt.x.org id KAA04374; Thu, 12 Oct 1995 10:26:48 -0400
Message-Id: <199510121426.KAA04374@exalt.x.org>
To: asami@cs.berkeley.edu (Satoshi Asami)
Cc: hackers@freefall.FreeBSD.org
Subject: Re: xload dumps core with new phkmalloc 
In-Reply-To: Your message of Wed, 11 Oct 1995 23:24:58 EST.
             <199510120624.XAA01000@silvia.HIP.Berkeley.EDU> 
Organization: X Consortium
Date: Thu, 12 Oct 1995 10:26:48 EST
From: "Kaleb S. KEITHLEY" <kaleb@x.org>
Sender: owner-hackers@FreeBSD.org
Precedence: bulk


Satoshi Asami said:

> Hello.  I found that xload dumps core with "floating point exception"
> with the latest phkmalloc (supped yesterday).  "MALLOC_OPTIONS=Z" will
> fix it, so it's definitely xload's problem.
> 
> I changed the only "malloc" in the xload source to "calloc" but that
> didn't help, so I guess the bug is in a more fundamental level.  This
> is what gdb has to say:
> 
> ===
> Program received signal SIGFPE, Arithmetic exception.
> 0x8034655 in repaint_window ()
> (gdb) bt
> #0  0x8034655 in repaint_window ()
> #1  0x8034284 in Redisplay ()
> #2  0x807992b in SendExposureEvent ()
> #3  0x80796f1 in CompressExposures ()
> #4  0x807939d in XtDispatchEventToWidget ()
> #5  0x8079df2 in L404 ()
> #6  0x807a128 in L458 ()
> #7  0x807a4b9 in XtAppMainLoop ()
> #8  0x1f75 in main (argc=1, argv=0xefbfcf10) at xload.c:264
> ===
> 

Poul-Henning Kamp said:

> Seems like you have to start looking at X11 then :-(

What does MALLOC_OPTIONS=Z mean? What does MALLOC_OPTIONS!=Z mean?
Does phkmalloc of zero bytes return NULL as it does in SVR4 when you 
link with -lmalloc, or on Linux, or or SunOS when you link with -llwp?

Dunno, if X works on dozens of platforms, with all the various and sundry 
mallocs that are out there like gnumalloc, bsdmalloc, Caltech malloc, 
Connor Cahill's malloc, Purify, Sentinel, etc., etc., but it doesn't work 
with phkmalloc, how daring would it be for me to suggest that phkmalloc 
is the culprit??? :-)

I don't want to have to figure out sup to get the source to phkmalloc,
I've just got too many other things on my plate at the moment. Is it on 
freefall in -current yet? Otherwise can someone tar up the sources,
uuencode them, and mail them to me?

--

Kaleb KEITHLEY
X Consortium