Skip to contentSkip to conversation

PROBLEM: 2.4.4ac7 oops, locks in init on boot

DISCUSSION Gordon Sadler 4 participants 7 messages May 11, 2001 – May 12, 2001

lore
Share & tools
lore
#0 lore raw

The ongoing saga of newer via chipsets and AMD CPU's. Please CC replies as I'm I read here via nntp.

Please see my previous PROBLEM reports for hardware info.

I compiled and boot 2.4.4ac7, got as far as init. Multiple oops's scrolled off the screen before I could see them. Attached is the final oops, but I have a feeling this is useless. You'll find my config and my attempts to run this oops through ksymoops at various -d levels. Ksymoops never returns from fnmatch or re_match_2, I had to ^C to get a prompt back. I would guess that has to do with eip: 00000000 ... Code:
Bad EIP value!

HTH

If anyone has any further suggestions/patches to run 2.4.x with K7 chosen optimizations, I'm open to testing.

Gordon Sadler

#1 lore raw
Quoted 2 lines

If anyone has any further suggestions/patches to run 2.4.x with K7 chosen optimizations, I'm open to testing.

'Buy an AMD chipset box..'

Seriously at this point I am out of ideas. The prefetch to far effect explained the old athlon locks (step 1) people reported on all chipsets. It didnt really seem to explain the problem with a few via chipset boards but I was hopeful.

Alan

#2 lore raw
Quoted 9 lines · Alan Cox
Quoted 2 lines

If anyone has any further suggestions/patches to run 2.4.x with K7 chosen optimizations, I'm open to testing.

'Buy an AMD chipset box..'

Seriously at this point I am out of ideas. The prefetch to far effect explained the old athlon locks (step 1) people reported on all chipsets. It didnt really seem to explain the problem with a few via chipset boards but I was hopeful.

So are you saying that given the current information available you have you do not know how to fix the via -> Athlon stuff or am I reading too much into this?

I am looking at buying an Athlon board soon and I am trying to figure out if I should avoid via at all costs or not.

TIA,

Signature · 3 lines
......Tom		DISSERVICE: It Takes Months to Find a Customer, but
[email protected]		Only Seconds to Lose One... The good News is that We
			Should Run Out of Them In No Time.
#3 lore raw
Quoted 3 lines

So are you saying that given the current information available you have you do not know how to fix the via -> Athlon stuff or am I reading too much into this?

I dont know how to fix it. Im not too sure I know how to even chase it down further. Im still curious about the way it only affects some people and don't know if its hw/sw either.

#4 lore raw
In reply to Gordon Sadler
Quoted 9 lines · Alan Cox
Quoted 2 lines

If anyone has any further suggestions/patches to run 2.4.x with K7 chosen optimizations, I'm open to testing.

'Buy an AMD chipset box..'

Seriously at this point I am out of ideas. The prefetch to far effect explained the old athlon locks (step 1) people reported on all chipsets. It didnt really seem to explain the problem with a few via chipset boards but I was hopeful.

I'm considering it, but for AGP weirdness reasons. Have the USB bugs been worked out? I am highly dependant on USB.

Signature · 4 lines
Joseph Carter <[email protected]>                Free software developer

<Palisade> how are we going to pronounce '00 or '01 or '02 and so on?
<Deek> Say goodbye to the nineties, say hello to the naughties. :)
#5 lore raw
Quoted 2 lines

I'm considering it, but for AGP weirdness reasons. Have the USB bugs been worked out? I am highly dependant on USB.

Only one stepping of the AMD chip had the USB funnies. AMD released the info needed to work around it and its now in the tree.

Alan

#6 lore raw
In reply to Joseph Carter
Quoted 5 lines · Alan Cox
Quoted 2 lines

I'm considering it, but for AGP weirdness reasons. Have the USB bugs been worked out? I am highly dependant on USB.

Only one stepping of the AMD chip had the USB funnies. AMD released the info needed to work around it and its now in the tree.

Ah, thanks Alan. I may need to consider the AMD boards seriously if I can't get the Radeon 64 working on the KT7A in the next few weeks. I'm hoping -ac7 will fix that.

Signature · 5 lines
Joseph Carter <[email protected]>                Free software developer

<taniwha> Knghtbrd: we should do a quake episode :knee deep in the code":
          you run around shooting at bugs:)
<Knghtbrd> taniwha: I'll pass the idea on to OpenQuartz  ;>