On Wed, Jul 31, 2002 at 02:21:22PM -0500, Hundstad, Jeffrey E. wrote:
> As requested I checked my facts. I'm not sure from the comments (included
> below) which AMD processor(s) have the problems. Sometimes they talk about
> newer AMD Athlon and sometimes its B step AMD K6 before B 9730xxxx. I also
> know that the XFS code doesn't use the AGP bus but a generic memory
> corruption in the cache caused by AGP use could corrupt kernel space used by
> XFS. True?
Not in this case. You can only get memory corruption on pages mapped into
the AGP aperture. While XFS can get later a page that was previously
mapped into the aperture it would not care about its previous (corrupted
or not) contents.
Also note the corruption is extremly rare and you are unlikly to trigger
it unless you beat the system heavily for many hours/days.
> * B step AMD K6 before B 9730xxxx have hardware bugs that can cause
> * misexecution of code under Linux. Owners of such processors should
> * contact AMD for precise details and a CPU swap.
This refers to a different bug.