prefetchnta on athlon was: Re: an oops in 2.6-test2 (oops)

From: Mike Fedyk
Date: Sun Aug 17 2003 - 11:52:23 EST


On Sat, Aug 16, 2003 at 12:26:40PM -0700, Andrew Morton wrote:
> And you'll continue to get them until someone does something about it.
> Discussion seemed to die off on this problem.
>
> Until it is sorted, something like this is needed.
>
>
> diff -puN include/asm-i386/processor.h~disble-athlon-prefetch include/asm-i386/processor.h
> --- 25/include/asm-i386/processor.h~disble-athlon-prefetch 2003-08-16 12:22:32.000000000 -0700
> +++ 25-akpm/include/asm-i386/processor.h 2003-08-16 12:23:29.000000000 -0700
> @@ -568,6 +568,8 @@ static inline void rep_nop(void)
> #define ARCH_HAS_PREFETCH
> extern inline void prefetch(const void *x)
> {
> + if (current_cpu_data.x86_vendor == X86_VENDOR_AMD)
> + return;

Andrew, if you put this patch in -mm please add a nice big comment above it,
and put it in the "must fix before 2.6.0" list.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/