Re: Please help - kernel crashes often

[lsorense@csclub.uwaterloo.ca (Lennart Sorensen)]

On Tue, Jan 31, 2006 at 08:09:16PM -0800, mike wrote:
> Do you mean disable the CPU cache (L2?) or disable the ECC memory cache?

I meant L2 CPU cache.

> I haven't looked at the BIOS yet to see which options I have, but I
> want to make sure I do the right one :)
> 
> Wouldn't disabling some caches wind up cutting the performance down,
> especially if you mean the CPU cache? Having the dual 1M L2 caches is
> supposed to be really good, disabling those would probably be a big
> performance hit then wouldn't it?

It will seriously hurt performance.  I don't think it is a solution,
just a debuging step to determine where the problem is.  After all if it
runs without errors with the cache turned off, and has errors with cache
turned on, it makes you suspect the cpu cache as the issue.  If the
system has more than one cpu you could always remove one cpu and see if
the system runs fine then, and then try with the other cpu.  Most likely
it would be only one of the cpus.  Or you could take out half the ram,
and see and try with the other half of the ram to see if you have a bad
stick of memory.

> Thanks for the feedback though. All of this is helping me.

Well the ECC errors certainly point to either a CPU cache problem or a
memory problem.  Problem is just finding out which one it is.  Of course
it could also be a motherboard problem, but then I would expect it to be
more frequent.

Len Sorensen