From owner-freebsd-questions  Fri Jun 28 10:02:51 1996
Return-Path: owner-questions
Received: (from root@localhost)
          by freefall.freebsd.org (8.7.5/8.7.3) id KAA14152
          for questions-outgoing; Fri, 28 Jun 1996 10:02:51 -0700 (PDT)
Received: from diablo.ppp.de (diablo.ppp.de [193.141.101.34])
          by freefall.freebsd.org (8.7.5/8.7.3) with SMTP id KAA14145
          for <questions@FreeBSD.org>; Fri, 28 Jun 1996 10:02:44 -0700 (PDT)
Received: from allegro.lemis.de by diablo.ppp.de with smtp
	(Smail3.1.28.1 #1) id m0uZeAy-000Qd0C; Fri, 28 Jun 96 16:05 MET DST
From: grog@lemis.de (Greg Lehey)
Organisation: LEMIS, Schellnhausen 2, 36325 Feldatal, Germany
Phone: +49-6637-919123
Fax:   +49-6637-919122
Received: (grog@localhost) by allegro.lemis.de (8.6.9/8.6.9) 
	id PAA12957; Fri, 28 Jun 1996 15:21:41 +0200
Message-Id: <199606281321.PAA12957@allegro.lemis.de>
Subject: Re: Compiling kernel for ATAPI CD-ROM error - fatal signal 11- what is it?
To: ajohn@mail.bcpl.lib.md.us (Anil John)
Date: Fri, 28 Jun 1996 15:21:41 +0200 (MET DST)
Cc: questions@FreeBSD.org (FreeBSD Questions)
In-Reply-To: <01BB6462.E7F53500@ppp15.bcpl.lib.md.us> from "Anil John" at Jun 27, 96 07:57:42 pm
X-Mailer: ELM [version 2.4 PL23]
MIME-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 8bit
Sender: owner-questions@FreeBSD.org
X-Loop: FreeBSD.org
Precedence: bulk

Anil John writes:
>
> Greg Lehey[SMTP:grog@lemis.de] wrote:
>>
>> OK, you're not done yet.  So it's pretty certain that you have a cache
>> problem.  Now you need to find out whether it's a hardware problem or
>> incorrect configuration.  Many BIOSes have cache burst configuration
>> options (cache wait states, or burst configurations like 3-2-2-2 or
>> 2-1-1-1 or some such).  The cache is 128 bits wide, but the memory bus
>> is only 32 bits wide, so it needs 4 cycles to transfer a cache line.
>> The 3-2-2-2 means that the first transfer takes 3 bus cycles, and the
>> following ones only 2.  I'd guess that you have your cache set up for
>> too few cycles.
>>
>
> I know what you mean The above settings are in my BIOS on my other machine.
>  But I have an old Award BIOS on my system and I do not know what the
> corresponding items are.  The relevent options on it are (* marks what I
> had when I got errors):
>
> Decoupled Refresh:  	Enabled*/Disabled
> Relocate 256k/384k:		Enabled*/Disabled
> Video BIOS Cacheble:	Enabled/Disabled*
> System BIOS Cacheble:	Enabled/Disabled*
>
> External cache scheme:	Wr-Through*/Write-Back
> Combine Alter & Tag Bits:	Enabled*/Disabled
> Memory hole at 15Mb Addr:	Enabled/Disabled*
> Cache Timing Control:	Normal/Fast/Turbo*
> DRAM Timing Control:	Normal/Fast*
> Fast DRAM:			Enabled*/Disabled
> Burst Write:			Enabled/Disabled*
> CPU Write Back Cache:	Enabled*/Disabled
> P24T Cache Replace BLAST:	Enabled/Disabled*

OK, at a guess I'd change these ones:

> Cache Timing Control:	Normal*/Fast/Turbo
> DRAM Timing Control:	Normal*/Fast
> Fast DRAM:			Enabled/Disabled*

If that fixes the problem, it will also slightly slow down the machine
(probably not enough to be noticable).  If you really want to squeeze
out the last 0.5% performance, you can then gradually try speeding
things up.  The trouble is, you can easily get the machine into a
marginally unstable situation where it fails in strange ways only when
your back it turned.

Greg