On Monday September 23 2002 11:30 am, Charles A Edwards wrote:
> On Mon, 23 Sep 2002 08:55:48 -0500
>
> Tom Brinkman <[EMAIL PROTECTED]> wrote:
> > Interesting, but it directly contradicts the statement made by
> > 'Sean
> > Cleveland and Wayne Meritsky of AMD' (the kernel ML link I gave). I
> > can see why AMD could release a patch for W2k, even tho the bug
> > isn't in their cpu's. Also, I believe the Linux 2.4 kernel has
> > been patched
> >
> > too, for quite some time now. IIRC, shortly after Mdk 8.0 was
> > released.
>
> Doing a little more digging today I was able to find a more detailed
> explanation.
> Unfortunately it gives credence to your view rather than to mine.
>
> http://marc.theaimsgroup.com/?l=linux-kernel&m=102376926732464&w=2
>
> It is from June of his year and very interesting.
>
> At any rate for whatever reason the nopentium append does work for X
> stability on my systems.
> BTW all my hardware is from the AMD approved lists and the power
> supplies are all in excess specs.
>
>
> Charles
Try without nopentium, but set the aperature to 4 in bios? Still
lock-up? IE, (from your link), "The kernel bug is often exposed in
conjunction with use of the AGP Aperture on these platforms." Which is
why I suggested a 4mb aperature in my earlier post. Then try again
with the open source drivers. BTW, the AGP Aperature set to 4 is an
old overclockers trick when the PCI bus is way out'a spec due to raisin
the FSB too high. Degrades video performance somewhat, but I wonder if
'nopentium' doesn't also (?).
I've seen you mention you have a Soyo Dragon +, an AMD apprv'd
board. What aggravates the AGP issue, any OS, is PSU's. Specially with
high performance AGP video cards from nVidia or ATI. Watt's, advertised
specs aren't the important PSU specs. Manufacturer is, particularly
model no., as they're marketed under several brand names (eg, I've got
a Powerman 300w PSU in an Inwin case, but the model no. reveals it's
made by Sparkle, and is an AMD appr'vd PSU model). IOW's, all 300+ watt
PSU's vary in 'cleaness', stability of power. AMD takes the effort to
figure out suitable ones, Intel does also.
I'm curious tho why a kernel bandaid (nopentium) would fix a problem
that was reported as fixed in recent kernels (for the last year), or
stems from marginal hardware, and/or wattage hoggin AGP cards. I still
suspect the video drivers are involved, specially the popular closed
source proprietary ones. IOW's I still believe the AGP problem might
be user (choices, ie, driver) and hardware centric.
--
Tom Brinkman Corpus Christi, Texas
Want to buy your Pack or Services from MandrakeSoft?
Go to http://www.mandrakestore.com
