On Tue, 22 May 2001, Richard Earnshaw wrote:

> > What is the effect of an instruction here?
>
> The processor can prefetch-abort when it shouldn't.

Can that not be worked around in the prefetch-abort handler? (I assume
not, but I'm interested to know why)

> > Is the panic I'm seeing consistent with this?
>
> I've not seen this before, but it could well be a symptom of the
> underlying problem.  Normally, a program starts looping interminably.

Then it would seem odd that a panic would result, no?

Is it likely that compiling a new kernel will make any difference?

> > Which binary should I be looking to patch? cvs? Or
> > the kernel?
>
> CVS.  It mainly happens on applications that start to swap, but there's no
> guarantee that other (small) programs won't show similar problems.

I did try turning off swap Just In Case[tm] and it still crashed.
Certainly last night it was 100% repeatable.

I'll have to see if it's repeatable once I have a complete source tree.

> The only real fix for the SA-rev2 problems is to replace it with an
> SA-rev3 or later chip... ;-(

:( So no magic or intelligent guesswork possible to restart the ldm
instruction from a better alignment within the abort handler?

-- 
Gillette - the best a man can forget