Re: [PATCH 4/4] x86_64 ioapic: Improve the heuristics for when check_timer fails.

[Eric W. Biederman]

"Lu, Yinghai" <yinghai.lu@xxxxxxx> writes:

>
>>I have tested this on an Nvidia board that reports that apic 0 pin 2
>>works when it does not and this code successfully programs apic 0 pin 0
>>into ExtINT mode.
>
> For some CK804 boards, BIOS forget set to 8254 timer to apic0/pin2, and
> still leave it at apic0/pin0. but mptable and acpi said 8254 timer is to
> apic0/pin2. At such case we should try apic0/pin0 with INT mode instead
> of ExtINT mode.

Duh.  I forgot Nvidia provided a register like that, now I am
beginning to understand what is wrong.  Anyway we do try apic0 pin0 by
default on the Nvidia's because of a Nvidia specific work around.

So that doesn't invalidate the generic test.  I'm going to go dig
out what little information I have and see if I can stair at the
register definition.

Anyway that doesn't invalidate anything in this patch. 

Eric
-
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/