Paulo Marques wrote:
> Bartlomiej Zolnierkiewicz wrote:
>>
>> Perhaps you can try first -rc3 git snapshot (still a lot of stuff):
>> http://www.kernel.org/pub/linux/kernel/v2.6/snapshots/patch-2.6.12-rc3-git1.bz2 
> 
> I tried this version with the same results.
> 
> [...]
> so that first "if" statement translates to
> 
>  >     } else if (strncmp(str, "PCI", 3) == 0) {
> 
> which should be always valid for "PCI   ", and that message should never 
> appear :(
> 
> Well, sorry about the noise, I must dig deeper now...

Well, I've dug deeper now. It seems that having "acpi=ht" disables acpi 
in 2.6.12-rc4, but not on 2.6.12-rc3.

Looking at the code it seems that it should have been disabled even on 
2.6.12-rc3.

I've attached the dmesg from 2.6.12-rc3, 2.6.12-rc4 with "acpi=ht" 
(which fails to boot) and 2.6.12-rc4 without any "acpi=" on the command 
line.

It seems to be running smoothly now. I know now that I shouldn't have 
used "acpi=ht", but why did it work on previous kernels?

Even more, the kernel with "acpi=ht" seems out of control, like it has 
some memory overwrite problems. Symptoms include the error "Unknown 
bustype PCI    - ignoring", and a lot of other strange errors ("unknown 
bus type 0.", "PCI BIOS passed nonexistent PCI bus 0!", etc.).

-- 
Paulo Marques - www.grupopie.com

An expert is a person who has made all the mistakes that can be
made in a very narrow field.
Niels Bohr (1885 - 1962)