lkml.org 
[lkml]   [2005]   [Oct]   [25]   [last100]   RSS Feed
Views: [more markup]   [less markup]   [headers]   [forward]  
 
Messages in this thread
/
DateTue, 25 Oct 2005 13:45:10 -0700 (PDT)
FromLinus Torvalds <>
SubjectRe: Call for PIIX4 chipset testers

On Tue, 25 Oct 2005, Darren Salt wrote:
>
>   $ dmesg | grep PIIX4
>   PCI quirk: region 5000-503f claimed by PIIX4 ACPI
>   PCI quirk: region 4000-401f claimed by PIIX4 SMB
>   PIIX4 devres C PIO at 0100-0107
>   PIIX4 devres I PIO at 00e0-00e3
>   PIIX4 devres J PIO at 00f9-00fc

Ok, great. None of those quirks are very interesting per se (they're all 
in the old legacy region), but that 0xf9-0xfc one is interesting if only 
because it points out that I should be more careful about the mask/base 
logic.

That quirk is actually set up in a very special way, where it ignores 
address line #1, and devres J actually covers two PIO ports: 0xf9 and 0xfb 
(but not the one between). Odd, but it's documented hardware behaviour.

So the 0xf9-0xfc thing is strictly not right, and I'll modify my code a 
bit (probably to mark 0xf8-0xfb instead, which will be what we'd want to 
do for anything in the non-legacy region).

> Machine is a Compaq Armada M700; /proc/ioports & lspci output are attached.

Thanks, that lspci output was what I needed to verify the mask details.

The printouts look good. I wish somebody had a PIIX with a MMIO quirk just 
to verify that I got that case right too, but it's pretty unlikely that is 
ever used in practice.

		Linus
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-10-25 20:47    [from the cache]
©2003-2008