*BSD News Article 45417


Return to BSD News archive

Newsgroups: comp.unix.bsd.freebsd.misc
Path: sserve!newshost.anu.edu.au!harbinger.cc.monash.edu.au!simtel!news.sprintlink.net!howland.reston.ans.net!Germany.EU.net!nntp.gmd.de!news.rwth-aachen.de!news.rhrz.uni-bonn.de!news.uni-stuttgart.de!rz.uni-karlsruhe.de!thorin!brewhq.swb.de!news.th-darmstadt.de!hrz-ws11.hrz.uni-kassel.de!phase23!citylink.dinoex.sub.org!peter
From: peter@citylink.dinoex.sub.org (Peter Much)
Subject: Re: Slight flame from Linux user
Organization: Buero fuer Sektenforschung und Qualitaetspruefung in der Esoterik
Message-ID: <D9t6un.EqI@citylink.dinoex.sub.org>
References: <3ql3gd$je2@bell.maths.tcd.ie> <3qp02d$eqb@news1.best.com> <3qpfm0$76j@bell.maths.tcd.ie> <3r166m$fit@bonnie.tcd-dresden.de>
Date: Wed, 7 Jun 1995 15:25:34 GMT
Lines: 49

In article <3r166m$fit@bonnie.tcd-dresden.de>,
J Wunsch <joerg_wunsch@uriah.heep.sax.de> wrote:
>Timothy Murphy <tim@maths.tcd.ie> wrote:
>
>>>	Fatal signal 11 syndrome is almost always a HARDWARE PROBLEM - 
>...
>
>>While the general opinion (including mine) is that the problem
>>is hardware-based, a minority believe it is a problem with gcc.


That's _NO_ hardware-problem. The day before yesterday we installed
a Linux 1.2.8 onto one of our hardware-gurus' portable. The platform has
already run other OS', at least an older Linux, maybe some *BSDs too.

I witnessed such a strange sig-11 failure in the middle of a compiler run
happening unreproduceable by chance. I witnessed "ls" working one time,
reporting "segment violation" the next time, and working again the third
time. I witnessed a make no longer able to compile smail 3.1.28. And that
was only the first afternoon.

My old Linux 0.99.pl15f Installation seems much more stable and consistent.

(To say it clearly: I get strangeness from my NetBSD 1.0 Installation,
too. And I do not yet have FreeBSD in an installation where I can give
it real hard work - need a 3rd PC first;-))

>>Some people claim that it only occurs when using gcc 2.6.?,
>>and that reverting to 2.5.8 will solve it.
>>So it is conceivable that the problem might not arise with FreeBSD.
>
>I'm surprised about your conclusion.  I hope you are aware that
>FreeBSD is using just the same gcc you're blaming here?

The linux kernel is loosing interrupts. This has been the case since at
least 0.99.pl6, and I do not think it has been corrected. Maybe in it's
root it is indeed a hardware problem coming from some misconception of
the pc mainboard's interrupt handling, but we had many boards under
test in the meantime...

So, I'm not surprized about such a conclusion. Linux folks were telling
all the time that these are hardware problems. Maybe they are, but
then they are on 11 of 12 boards.


Peter
-- 
  Write to:  Peter Much * Koelnische Str. 22 * D-34117 Kassel * +49-561-774961
 peter@citylink.dinoex.sub.org  much@hrz.uni-kassel.de   p.much@asco.nev.sub.de