Re: [ACPI] Strange speedstep-related code in some DSDT

[Bruno D. <du...@po...>]

On Tue, Nov 08, 2005 at 10:57:01AM +0100, Colin Leroy wrote:
> On 08 Nov 2005 at 10h11, Bruno Ducrot wrote:
> 
> Hi, 
> 
> > On Wed, Nov 02, 2005 at 03:00:10PM +0100, Colin Leroy wrote:
> > > Hi,
> > > 
> > > After I upgraded the BIOS on an Asus M6B600C laptop, I found it to
> > > be stuck at 600MHz. After investigating for a while using
> > > debug.cpufreq=3, I found out that the DSDT looked strange, and
> > > after I fixed it, cpufreq worked again.
> > > 
> > > The things I had to fix were mainly:
> > > -            Name (APCT, Package (0x02)
> > > +            Name (_PCT, Package (0x02)
> > > 
> > > and
> > > -            Name (NPSS, 0x00)
> > > -            Name (APSS, Package (0x0A)
> > > +            Name (NPSS, 0x05)
> > > +            Name (_PSS, Package (0x0A)
> > > 
> > > I am wondering whether such objects should be recognized, or is it
> > > Asus' fault? I read somewhere that they ship a (windows-based)
> > > proprietary speed controller software for their laptops, so they
> > > probably don't care about specs in this regard.
> > > 
> > > Do you think a patch would be ok to handle these in
> > > processor_perflib.c? 
> > > 
> > 
> > I'll suspect more that after upgrading your bios you forgot to enable
> > or disable speedstep stuff based upon a bios option...
> 
> Thanks for the answer. I thought of that, but it is true that everytime
> I got back in the bios, speedstep was flagged 'disabled'. Maybe a bug
> BIOS then :)

Most likely.

But if a propritary driver work, I think the speedstep-smi may work
since it's actually more or less a reverse-engineered driver...

-- 
Bruno Ducrot

--  Which is worse:  ignorance or apathy?
--  Don't know.  Don't care.