Re: Compiling problems 1.3.85

David A Willmore (willmore@cig.mot.com)
Tue, 9 Apr 1996 14:45:18 -0500

On Apr 9, 10:34am, tim middlekoop wrote:
> Subject: Re: Compiling problems 1.3.85
> I am starting to beleve that these signal 11 problems are a result of AMD
> cpu's and a quicker memcpy. Many people suggest that disabling the
> secondary cash solves signal 11 problems, It does for me. Possibly
> disabling the secondary cache has the same problem solving effect as the
> slower memcpy.
> tim...

Disabling the L2 cache on 486 motherboards does change the memory timing to be
a bit more relaxed. Your real problem, most likely, is that you are running
the memory too fast. I can run my 486DX4/120 with either fast memory timing
and one wait state or slow memory timing and no wait states. The latter
benchmarked slightly faster and I haven't had a Sig 11 since I made the change.

The BogoMIPS number is still messed up. I wish I had an explanation for that
one.

Cheers,
David