Xref:	news2.mv.net comp.os.msdos.djgpp:2669
From:	Broeker AT axp03 DOT physik DOT rwth-aachen DOT de (Hans-Bernhard Broeker)
Newsgroups:	comp.os.msdos.djgpp
Subject:	Re: malloc crash
Date:	11 Apr 96 13:32:30 GMT
Organization:	RWTH -Aachen / Rechnerbetrieb Informatik
Lines:	18
Message-ID:	<Broeker.829229550@axp03>
References:	<199604100421 DOT AA172870100 AT relay1 DOT geis DOT com> <316B7A47 DOT 381F AT matrust DOT es> <4kicd3$njj AT freenet-news DOT carleton DOT ca>
NNTP-Posting-Host:	axp03.physik.rwth-aachen.de
To:	djgpp AT delorie DOT com
DJ-Gateway:	from newsgroup comp.os.msdos.djgpp

ao950 AT FreeNet DOT Carleton DOT CA (Paul Derbyshire) writes:

>I saw this happen too.
[...]
>...causing it to
>allocate 10000 bytes every line. Then unawares I had it read in a large
>file and it sat "thinking" for a while then crashed with a GPF of some
>kind or another. 

Well, I'm not sure if this is really related, but some weeks ago I had
problems to compile one specific source file (part of gnuplot 3.6) with
"-O3 -Wall -W" flags enabled (cc1 produced a seg fault).  After some
tracking down, I finally managed to 'solve' this by rebuilding gcc and
using the GNU malloc package (the offending line inside cc1 was actually
a call to malloc, and that error vanished after the switch to GNU's
malloc).  Maybe either cwsdpmi or DJ's malloc is buggy?

Hans-Bernhard Broeker (Aachen, Germany)

- Raw text -