Re: [OT][Wanted] 667MHz Slot-B Alpha CPU

[Marc Zyngier <maz misterjones org>]

On Sat, 16 Aug 2008 17:47:50 -0600
Maurice Hilarius <maurice harddata com> wrote:

> Robert M. Riches Jr. wrote:
> >> Date: Sat, 16 Aug 2008 15:15:27 -0600
> >> From: Maurice Hilarius <maurice harddata com>
> >>
> >> Marc Zyngier wrote:
> >>     
> >>> Folks,
> >>>
> >>> It looks like one of the CPU on my UP-2000+ gave up the ghost
> >>> (machine stopped in the middle of a compilation). I took away
> >>> every single part of the machine and it works fine, except when I
> >>> insert the second CPU. Sad day...
> >>>
> >> Or, it could be the UP-2000+ motherboard..
> >>
> >> Just so you know this can sometimes show the symptom you described.
> >>     
> >
> > Would swapping slots be the best method to diagnose whether
> > the problem is with the CPU or the motherboard?
> >
> >   
> Pretty much.

At first, I suspected the board, given its relative unreliability track
record.

But then I tried that:
- put CPU "A" in slot 1, CPU "B" in slot 2, doesn't work
- put CPU "A" in slot 2, CPU "B" in slot 1, doesn't work
- put only CPU "A" in slot 1, works
- put only CPU "B" in slot 1, doesn't work
- put only CPU "A" in slot 2, works (surprisingly !)
- put only CPU "B" in slot 2, doesn't work

In dual CPU configuration, "doesn't work" symptom is either SRM hanging
just after memory probing, or going in an endless stream of register
dump after an access fault.

In single CPU configuration, "doesn't work" symptom is that the SROM
LED stays lit, without any console output at all.

This is why I strongly suspect CPU "B".

	M.
-- 
And if you don't know where you're going, any road will take you there.