K7/Athlon optimizations again. (The sacrifices worked??) (VIA KT133A chipset)

K7/Athlon optimizations again. (The sacrifices worked??) (VIA KT133A chipset)

[Nicholas Knight]

Thanks to Charles Cazabon, I now have a new theory.
to quote, I'm the one he quoted here:

----Begin Quote---
> And I CAN nail it down to a particular chipset. It only occurs on VIA 
> KT133A chipsets so far, unfortunitely it doesn't occur 100% of the time 
> on that chipset.


I'd be curious if it's an issue with only one brand or release of BIOS
then.
----End Quote----

And thus begins a new round, I started by slapping myself.

I've seen two instances so far of someone mentioning that one BIOS 
release caused problems while another did not. One was mentioned on the 
list, one was part of the reports sent to me. I completely passed over 
this information and I have no idea why.

Everyone having problems, here's what you could try:
Upgrade to the latest BIOS version for your board.
If you are at the latest BIOS version, go back a version or two.

For those NOT having problems, and who would like to help test/live 
dangerously, could you also try the previously mentioned upgrades or 
downgrades? 

And be sure to have a non-K7 optimized kernel installed as a backup 
incase you forget to make TWO diskettes for upgrading your BIOS and it's 
your only system :)

This problem doesn't even affect me, I'm not a developer, and any system 
I buy in the future will be long after the KT133A is "outdated", I don't 
know why I'm hammering at this so much.

Re: K7/Athlon optimizations again. (The sacrifices worked??) (VIA KT133A chipset)

[Gerbrand van der Zouw]

Nicholas Knight wrote:

> I'd be curious if it's an issue with only one brand or release of BIOS
> then.


Hi,

I am one of the person reporting success with some BIOS version over 
another. I have been experimenting some more and found the following:

Sytem:
Board: MSI K7T Turbo (MS-6330)
CPU: Athlon 1.2 GHz
Video: NVidia GF2
Kernel: 2.4.6ac5 with the patch Kurt Garloff posted here some while ago 
(can't find the right reference)

Ranking from very unstable to stable:
Very unstable: 2.4.6 plain, Award V2.7 Bios (i.e. without the 
Southbridge bug solved the official VIA way)
Unstable: 2.4.6ac5 with or without Kurt's patch, Award V2.7 Bios
Stable (up to now): 2.4.6ac5 with Kurt's patch and Award V2.8 Bios.

The key factor seemed to be the BIOS. The MSI website says that the 
following things changed between release 2.7 and 2.8:
- Fix STR Fail on MS-6330Lite.
- Fix 3Dmark 2001 sometimes halt
- Fix K7T Turbo Limited cannot adjust vcore

Only the second fix seems applicable to my mobo, but the description 
does not give a clue.

I hope that someone can distill something useful out of this. I am of 
course willing to try some other combination of parameters if this would 
help finding the source of the problem.


Cheers,

Gerbrand van der Zouw

Re: K7/Athlon optimizations again. (The sacrifices worked??) (VIA KT133A chipset)

[Luigi Genoni]

On KT133A based MBs, I came trought at less 4 bios upgrades, and no
problems at all. Usually I do not oveclock CPUs, so i keep 1.725 volts for
vcore instead of 1.75, and system are just STABLE. In the reality setting
1.725 volts in the bios gives to my CPUs on abit MBs the right 1.75 volts,
and setting 1.75 gives them 1.78/9. So I would suggest also to investigate
if this light overvolt could be involved to create instabilities.
I just made a test, and saw that on athlon 1.33 Ghz this volts diffecence
creates a 5 degres difference when CPU is idle, and 7/8 when it is full
loaded. And belive me, actually i am keeping thos CPU in a 18 degrees
environment, using delta's fan to dissipate the eath.

I know that a light overvolt should increase stability, and not generate
instability, but maybe if CPUs and north bridge get too hot, (not
so hot to destroy the HW, but anyway hotter in front of the optimal
temperature), that could generate instability on loaded systems.

just my 2 cents

Luigi



On Thu, 6 Sep 2001, Nicholas Knight wrote:

> Thanks to Charles Cazabon, I now have a new theory.
> to quote, I'm the one he quoted here:
>
> ----Begin Quote---
> > And I CAN nail it down to a particular chipset. It only occurs on VIA
> > KT133A chipsets so far, unfortunitely it doesn't occur 100% of the time
> > on that chipset.
>
>
> I'd be curious if it's an issue with only one brand or release of BIOS
> then.
> ----End Quote----
>
> And thus begins a new round, I started by slapping myself.
>
> I've seen two instances so far of someone mentioning that one BIOS
> release caused problems while another did not. One was mentioned on the
> list, one was part of the reports sent to me. I completely passed over
> this information and I have no idea why.
>
> Everyone having problems, here's what you could try:
> Upgrade to the latest BIOS version for your board.
> If you are at the latest BIOS version, go back a version or two.
>
> For those NOT having problems, and who would like to help test/live
> dangerously, could you also try the previously mentioned upgrades or
> downgrades?
>
> And be sure to have a non-K7 optimized kernel installed as a backup
> incase you forget to make TWO diskettes for upgrading your BIOS and it's
> your only system :)
>
> This problem doesn't even affect me, I'm not a developer, and any system
> I buy in the future will be long after the KT133A is "outdated", I don't
> know why I'm hammering at this so much.
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at  http://www.tux.org/lkml/
>

Re[2]: K7/Athlon optimizations again. (The sacrifices worked??) (VIA KT133A chipset)

[victor]

Hello Luigi,

Friday, September 07, 2001, 12:01:50 AM, you wrote:

LG> On KT133A based MBs, I came trought at less 4 bios upgrades, and no
LG> problems at all. Usually I do not oveclock CPUs, so i keep 1.725 volts for
LG> vcore instead of 1.75, and system are just STABLE. In the reality setting
LG> 1.725 volts in the bios gives to my CPUs on abit MBs the right 1.75 volts,
LG> and setting 1.75 gives them 1.78/9. So I would suggest also to investigate
LG> if this light overvolt could be involved to create instabilities.
LG> I just made a test, and saw that on athlon 1.33 Ghz this volts diffecence
LG> creates a 5 degres difference when CPU is idle, and 7/8 when it is full
LG> loaded. And belive me, actually i am keeping thos CPU in a 18 degrees
LG> environment, using delta's fan to dissipate the eath.

look at this
http://bp6.gamesquad.net/Q6fix.phtml
i have a bp6 who has the same problem whit the dual celeron, and here
is the solution for it, looks your motherboard and see if its the same
problem, the celerons are fine whit this overvolt, they are prety
stable

its abit ... it cuold be ;)

LG> I know that a light overvolt should increase stability, and not generate
LG> instability, but maybe if CPUs and north bridge get too hot, (not
LG> so hot to destroy the HW, but anyway hotter in front of the optimal
LG> temperature), that could generate instability on loaded systems.

LG> just my 2 cents

LG> Luigi

-- 
Best regards,
 victor                            mailto:ixnay@infonegocio.com

Re: K7/Athlon optimizations again. (The sacrifices worked??) (VIA KT133A chipset)

[brian]

I have 10 Athlon/Duron based machines under my control.

Among the motherboards are:

Iwill KK266
Epox 8KTA2L
Epox 8KTA3+
FIC AZ11

They have various speeds of Athlons and Durons.  All are
standard (not overclocked) and have lots of cooling.

They all have oopses when trying to boot 2.4.9ac5 with Athlon/Duron
selected as the processor type.  They all work fine with  
6x86 or K6/K6-2/K6-3 selected as processor type.  (when not
running 2.4, they are running 2.2.19, and will do so for
months at a time).

Previously all the machines were in use so I couldn't offer
much assistance.  However, an Epox 8KTA2L/Duron 900MHz is
not being used at the moment, so if someone wants me to
try a patch or something I'd be happy too.

-- 
Brian Litzinger <brian@worldcontrol.com>

    Copyright (c) 2000 By Brian Litzinger, All Rights Reserved

Re: K7/Athlon optimizations again. (The sacrifices worked??) (VIA KT133A chipset)

[Jonathan Morton]

>Everyone having problems, here's what you could try:
>Upgrade to the latest BIOS version for your board.
>If you are at the latest BIOS version, go back a version or two.
>
>For those NOT having problems, and who would like to help test/live
>dangerously, could you also try the previously mentioned upgrades or
>downgrades?

I'd love to try the latest BIOS (I'm currently using a no-problems 
machine with it's original BIOS) but...  the IDE disk containing my 
root partition just appeared to go belly-up in the past hour or so. 
UGH!
-- 
--------------------------------------------------------------
from:     Jonathan "Chromatix" Morton
mail:     chromi@cyberspace.org  (not for attachments)
website:  http://www.chromatix.uklinux.net/vnc/
geekcode: GCS$/E dpu(!) s:- a20 C+++ UL++ P L+++ E W+ N- o? K? w--- O-- M++$
           V? PS PE- Y+ PGP++ t- 5- X- R !tv b++ DI+++ D G e+ h+ r++ y+(*)
tagline:  The key to knowledge is not to rely on people to teach you it.