Re: Latest on Athlon Via KT133A chipset solution?

Re: Latest on Athlon Via KT133A chipset solution?

[Seth Goldberg]

Hi,

I noticed discussion of the various Via KT133A chipset problems related
to 
Athlon optimized kernels has trailed off. Are people successfully using 
the patch that Alan Cox posted, or are there still problems? 

I just ran into this last night (I thought all the Athlon chipset bugs
had 
been fixed in 2.4.4 prior to last night). 

Anyway, just requesting status, and I'll gladly offer any testing help 
that's needed. 

====

  Actually it appears that I cut cut off from the kernel list :).  Hmm
did I say something wrong? :).  Seriously, there has not been any change
by me... I'm running 2.4.4 with the athlon-specific fast_page_copy
TURNED OFF.
My system is very stable, though I'm sad that I can't get the
performance
out of my system as I should... I guess it may be time to look for
another
M/B that does not exhibit these problems...

 --Seth (bergsoft@home.com)

Latest on Athlon Via KT133A chipset solution?

[kernel]

I noticed discussion of the various Via KT133A chipset problems related to
Athlon optimized kernels has trailed off. Are people successfully using
the patch that Alan Cox posted, or are there still problems?

I just ran into this last night (I thought all the Athlon chipset bugs had
been fixed in 2.4.4 prior to last night).

Anyway, just requesting status, and I'll gladly offer any testing help
that's needed.

Re: Latest on Athlon Via KT133A chipset solution?

[Alan Cox]

> I just ran into this last night (I thought all the Athlon chipset bugs had
> been fixed in 2.4.4 prior to last night).

Nope..

> Anyway, just requesting status, and I'll gladly offer any testing help
> that's needed.

Give the current -ac a spin and tell me if it works/doesnt and if not how
it fails

Re: Latest on Athlon Via KT133A chipset solution?

[Christian Bornträger]

> Give the current -ac a spin and tell me if it works/doesnt and if not how
> it fails

I will test it on Sunday evening, when I get back access to my Athlon.

Again I am not sure if I had the same problem related to the
compiler-optimizations. My problems (system freeze) started with 2.4.3-ac7.
So the problem might related to the VIA-bug-workaround.

Is there a possibility to get debugging messages or register dumps without a
second PC? Or is there a possibility to an unbuffered log write? e.g. write
kernel logs to a floppy disc unbuffered. Otherwise I am not sure, if I am
able to help you finding the problem.
Unfortunately I have only a VIA-based Athlon system and not a S/390

greetings

Christian Bornträger

Re: Latest on Athlon Via KT133A chipset solution?

[Alan Cox]

> Is there a possibility to get debugging messages or register dumps without a
> second PC? Or is there a possibility to an unbuffered log write? e.g. write

Not much.  You can avoid running syslogk. The one unbuffered log source
is the screen in text mode if not running it via syslogk

Re: Latest on Athlon Via KT133A chipset solution?

[Jeremy]

Just tested the -ac7 with K7 optimizations on.  I see the same behavior of
oopsing right after boot:

[...]
Freeing unused kernel memory: 216k freed
Unable to handle kernel NULL pointer dereference at virtual address
00000000
[...]

Just fine when I go back down to k6-3, et al, optimizations.

Tyan S2380B VIA KT133A, amd 1.33ghz, regular stepping, no OC or anything.

If you want more info, lemme know.

Also had some USB weirdness, but I'm going to debug that a bit before I
report on it.

thx,
-jeremy

Christian Bornträger enlightened recipients with the following on 11May2001:
> > Give the current -ac a spin and tell me if it works/doesnt and if not how
> > it fails
> 
> I will test it on Sunday evening, when I get back access to my Athlon.
> 
> Again I am not sure if I had the same problem related to the
> compiler-optimizations. My problems (system freeze) started with 2.4.3-ac7.
> So the problem might related to the VIA-bug-workaround.
> 
> Is there a possibility to get debugging messages or register dumps without a
> second PC? Or is there a possibility to an unbuffered log write? e.g. write
> kernel logs to a floppy disc unbuffered. Otherwise I am not sure, if I am
> able to help you finding the problem.
> Unfortunately I have only a VIA-based Athlon system and not a S/390

Re: Latest on Athlon Via KT133A chipset solution?

[Christian Bornträger]

 Give the current -ac a spin and tell me if it works/doesnt and if not how
> it fails

I tried 2.4.4-ac8.
It still hangs during boot if compiled for Athlon, but also crashes after a
while if compiled for 586. (I did a bonnie, did it a second time. The second
run doesn´t finished.) It is a complete system freeze, magic sysrq doesn´t
work.

Now i am going to try the new bios with a new version of the via fix. I will
report if this solution fix it for me/us.

greetings

Christian

Re: Latest on Athlon Via KT133A chipset solution?

[kernel]

Just to chime in since I restarted this thread. Tried 2.4.4-ac8 /wo
success, the same set of oops's that everyone else gets. Oh well, I
appreciate the attempt.

On Fri, 11 May 2001, Alan Cox wrote:

> > I just ran into this last night (I thought all the Athlon chipset bugs had
> > been fixed in 2.4.4 prior to last night).
>
> Nope..
>
> > Anyway, just requesting status, and I'll gladly offer any testing help
> > that's needed.
>
> Give the current -ac a spin and tell me if it works/doesnt and if not how
> it fails
>