On Sun, Nov 02, 2003 at 12:52:03PM +0000, Dave Jones wrote:
> On Sun, Nov 02, 2003 at 01:57:48PM +0800, Geoffrey Lee wrote:
> 
>  >  	preempt_disable(); 
>  > +#if CONFIG_MK7
>  > +	for (i=1; i<nr_mce_banks; i++) {
>  > +#else
>  >  	for (i=0; i<nr_mce_banks; i++) {
>  > +#endif
>  >  		rdmsr (MSR_IA32_MC0_STATUS+i*4, low, high);
> 
> This needs to be a runtime check. In 2.6, a K7 can boot
> a P4 kernel, and vice versa.
> 


(Resending as it seems to have eaten my mail due to dns problems ... apologies
if you get this twice.)


Would checking boot_cpu_data.x86_vendor == X86_VENDOR_AMD and 
boot_cpu_data.x86 == 6 be sufficient?  It seems to do the right thing ..
Updated patch attached.

	- g.