Hello Kent and BSD'ers from all sides !

- What's the model / stepping / revision of the incrimined CPU ? (un-dust
it, in case of !)

- Did you try to poke into BIOS for fail-safe settings, slow down RAM, I/O,
en/disable CPU caching, ECC, AGP aperture, etc...for debugging ?

- Sometimes, a low CMOS backup battery could make troubles.... I had the
case under NetBSD / VAX where the toy clock resetted 'hazardously', made a
kernel panic and kicked me into debbuger...

- Recently I've installed XP (a disk suffer of it...) and worked whell until
I flashed a newer BIOS release XP incompatible...$-/

- Did you ever try to boot M$-98 under the same configuration ? ( I've
noticed that it crash easily on boot if something bad with CPU / RAM ;-])

david.vyskocil@wanadoo.fr (please, don't blame my poor English...)


> I had pretty much determined it was a hardware error, but I ran
> memtest86 through a couple of complete cycles and it found nothing.
> Since there is only a gfx card, nic and ram in the box, and it
> works the same with and without the nic, it appears that the hardware
> failure is not triggered when one runs mem86.  I have re-inserted
> all of the cards and cables (and the cpu) in the system, which
> doesn't appear to affect anything.
>
> Interestingly, memtest86 runs with no problem.  When I can boot
> NetBSD, it will run until I do something as trivial as select
> another window under twm. Then it freezes and again pagefaults on
> boot. Again, memtest runs fine every time, but it appears that I
> can only boot NetBSD if I let the box remain powered off for some
> time. Is this starting to sound like the cpu is failing???
>
> Thanks