[GRLUG] Kernel panic?

[Tim Schmidt]

On Wed, May 14, 2008 at 4:02 PM, Michael Mol
> URL, to make sure you get the right one:
> http://www.memtest.org/
>
> There's another one, called Memtest86, but it's ancient and buggy.

In my experience, a large gcc compile job - like 'make allyesconfig'
in a fresh kernel tree - does a much better job coaxing memory errors
from faulty machines.  I've had machines pass multiple days of
memtest86 and memtest86+ without error, but repeatedly segfault during
such a compile.  The same compile job, on the same source tree, on the
same base OS install on different hardware worked fine.  Solution:
faulty memory.  It was replaced and all was well.

I don't trust memtest86 any more.

--tim