linux-mips
[Top] [All Lists]

Re: Segfaults on r4600

To: Ralf Baechle <ralf@oss.sgi.com>
Subject: Re: Segfaults on r4600
From: Guido Guenther <guido.guenther@gmx.net>
Date: Tue, 24 Jul 2001 10:24:37 +0200
Cc: linux-mips@oss.sgi.com
In-reply-to: <20010724052441.A25160@bacchus.dhis.org>; from ralf@oss.sgi.com on Tue, Jul 24, 2001 at 05:24:42AM +0200
Mail-followup-to: Ralf Baechle <ralf@oss.sgi.com>, linux-mips@oss.sgi.com
References: <20010723194830.A9033@galadriel.physik.uni-konstanz.de> <20010724052441.A25160@bacchus.dhis.org>
Sender: owner-linux-mips@oss.sgi.com
User-agent: Mutt/1.2.5i
On Tue, Jul 24, 2001 at 05:24:42AM +0200, Ralf Baechle wrote:
> On Mon, Jul 23, 2001 at 07:48:30PM +0200, Guido Guenther wrote:
> 
> > I'm seeing various segfaults especially with perl on an R4600 Indy.
> > R4000 I2 with identical debian packages works fine though. I have tried
> > various kernels (cvs head as from two days ago(natively and
> > crosscompiled) and 2.4.3-r4k-ip22 from rfc822.org. Interesting enough
> > the segfaults disappear when using "strace -o/dev/null
> > segfaulting_binary". I also tried to investigate the core file but gdb
> > dies when loading it(gdb is 5.0-3 debian package).
> 
> Strace uses ptrace(2) which does enormous numbers of cache flushes.
> Are those segfault non-deterministic?  Do the perl script you're running
> do alot of I/O?
No extensive I/O and perfectly deterministic. Another interesting fact
is, that some of these scripts fail when executed from a shells script 
but work perfectly well when exactly the same call is typed into a
shell.
 -- Guido

<Prev in Thread] Current Thread [Next in Thread>