2006-03-10 03:25:50

by Philip Langdale

[permalink] [raw]
Subject: pci-pci-quirk-for-asus-a8v-and-a8v-deluxe-motherboards.patch

Saw this patch mentioned in passing and I wanted to point out
that this is a more general problem than just for asus motherboards.
I have a soyo kt880 based dragon 2 motherboard and it exhibits the
same behaviour and the same fix works (modulo looking for a different
subsystem vendor - soyo apparently don't have their own id - it's
set to VIA for all devices). And I've read about it affecting other
motherboards - I think it must be something that's present in the
reference BIOS that all the manufacturers use.

I'm not sure what the most efficient way to generalise it - especially
with cases like the Soyo one where there's no proper subvendor id.

--phil


2006-03-10 06:41:33

by Greg KH

[permalink] [raw]
Subject: Re: pci-pci-quirk-for-asus-a8v-and-a8v-deluxe-motherboards.patch

On Thu, Mar 09, 2006 at 07:25:43PM -0800, Philip Langdale wrote:
> Saw this patch mentioned in passing and I wanted to point out
> that this is a more general problem than just for asus motherboards.
> I have a soyo kt880 based dragon 2 motherboard and it exhibits the
> same behaviour and the same fix works (modulo looking for a different
> subsystem vendor - soyo apparently don't have their own id - it's
> set to VIA for all devices). And I've read about it affecting other
> motherboards - I think it must be something that's present in the
> reference BIOS that all the manufacturers use.
>
> I'm not sure what the most efficient way to generalise it - especially
> with cases like the Soyo one where there's no proper subvendor id.

Great, thanks for the information. How about just adding new device ids
for the new machines that also need this function called. It's quite
easy to do that...

thanks,

greg k-h

2006-03-10 07:20:41

by Philip Langdale

[permalink] [raw]
Subject: Re: pci-pci-quirk-for-asus-a8v-and-a8v-deluxe-motherboards.patch

Greg KH wrote:
>>
>> I'm not sure what the most efficient way to generalise it - especially
>> with cases like the Soyo one where there's no proper subvendor id.
>
> Great, thanks for the information. How about just adding new device ids
> for the new machines that also need this function called. It's quite
> easy to do that...

Actually, I would go as far as to say just drop the check for the
subsystem vendor completely and unconditionally read the pci config
byte.

--phil