linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Roberto Jung Drebes <drebes@inf.ufrgs.br>
To: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: linux-kernel@vger.kernel.org
Subject: Re: Stomping on Athlon bug
Date: Thu, 13 Sep 2001 15:09:51 -0300 (EST)	[thread overview]
Message-ID: <Pine.GSO.4.21.0109131505170.25973-100000@jacui> (raw)
In-Reply-To: <m1r8tbt5i7.fsf@frodo.biederman.org>

On 13 Sep 2001, Eric W. Biederman wrote:

> Jan Niehusmann <jan@gondor.com> writes:
> 
> > On Thu, Sep 13, 2001 at 08:21:49AM -0400, Arjan van de Ven wrote:
> > > On Thu, Sep 13, 2001 at 02:19:38PM +0200, Jan Niehusmann wrote:
> > > > But, as far as I understand, STPGNT will not be enabled unless ACPI
> > > > power saving is in use, so setting the disconnect on STPGNT bit should
> > > > not matter.
> > > 
> > > That is incorrect; it works perferctly well without ACPI.
> > 
> > Exactly what is incorrect?
> > AFAICS, STPGNT is not triggered by hlt, so the linux idle function
> > doesn't set STPGNT.
> 
> Hmm.  At least on the AMD76[12] you can trigger a processor disconnect
> on hlt.  However the buggy BIOS had disconnects disabled so it doesn't/shouldn't
> matter.

Perhaps this is connected to this question of the unofficial KT7A
motherboard, available at http://www.viahardware.com/faq/kt7/kt7faq.htm:

Why is my idle temperature so much higher under BIOS version 3R or later?

With previous BIOS versions ABIT set a bit in the BIOS CMOS (offset 52 bit
7) to "1" to enable the ACPI "HALT" function and cool down the CPU temp in
idle mode.  This, whilst allowing cooler CPU operation, is against the
advice of AMD and VIA.  Under earlier AMD processors, however, this
modification allowed the system to idle cooler and had no ill effects.  
However, it was been found with newer 133MHz FSB processors running at
1333MHz and higher, that this software cooling can cause severe stability
problems - especially under Windows 2000.  As a consequence, in all ABIT
BIOS releases after version 3R, ABIT have set this bit to "0" - the value
recommended by AMD and VIA.  This allows the motherboard to be stable with
all AMD processors at the expense of an increase of 5-10 degrees
Centigrade during idle (from approximately 35 degs C to 45 degs C).  This
will not harm your motherboard or processor.  Note that this CMOS setting
is widely used by other KT133/KT133A based motherboards.  Note that if
your processor is slower than 1333MHz then you can revert to the previous
setting of the BIOS by using H-Oda's WCPREDIT and WCPRSET programs to
modify this register value.  If you are using Hex mode in these programs
then change register 52 from 6B to EB to re-enable the software cooling.  
See downloads page for programs.

??


--
Roberto Jung Drebes <drebes@inf.ufrgs.br>
Porto Alegre, RS - Brasil
http://www.inf.ufrgs.br/~drebes/


  reply	other threads:[~2001-09-13 18:09 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2001-09-13  9:13 Stomping on Athlon bug VDA
2001-09-13 10:06 ` Gergely Tamas
2001-09-13 11:27   ` Gergely Tamas
2001-09-13 10:17 ` Arjan van de Ven
2001-09-13 12:19   ` Jan Niehusmann
2001-09-13 12:21     ` Arjan van de Ven
2001-09-13 14:02       ` Jan Niehusmann
2001-09-13 16:33         ` Eric W. Biederman
2001-09-13 18:09           ` Roberto Jung Drebes [this message]
2001-09-13 11:58 ` Carsten Leonhardt
2001-09-13 15:11 ` Eric W. Biederman
2001-09-14 20:06   ` Pavel Machek

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=Pine.GSO.4.21.0109131505170.25973-100000@jacui \
    --to=drebes@inf.ufrgs.br \
    --cc=ebiederm@xmission.com \
    --cc=linux-kernel@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).