Re: Corruption bug in CPU, kernel or tar

Aaron J. Grier (agrier@poofy.goof.com)
Fri, 16 Jul 1999 08:49:07 -0700


On Wed, Jul 14, 1999 at 01:06:13AM -0700, Dan Hollis wrote:

> I have noticed that the motherboard voltage supplied to AMD K6 cpu by
> FIC motherboards is sometimes too low. Try upping the voltage a notch
> or two and see if that fixes it. It did for us on several K6 systems
> which were experiencing random corruption / sig11.

I'll have to try this on my K6/2-400 [1] at work. I've been suspecting
bad memory, but making the memory timings more aggresive didn't seem to
affect anything. I could do make -j20 -l10 on a full GNU tools compile
concurrently with a kernel compile and not have any problems. However,
all the times when I did get a sig11, I was running the GIMPS [2] mprime
program, which makes heavy use of the FPU.

I'll try upping my voltage a notch and let the list know what happens.

[1] with FIC PA-2013 motherboard, 1MB cache, 128MB of ECC SDRAM.
[2] the Great Internet Mersenne Prime Search (http://www.mersenne.org/)

----
Aaron J. Grier | "Not your ordinary poofy goof." | agrier@poofy.goof.com
"I really admire your perverse mastery of the SPARC branch delay slot,
Dave. Or is it your mastery of the perverse branch delay slot?"
-- Joe Martin to Dave S. Miller on linux-kernel

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/