cpufreq on athlon4

cpufreq on athlon4

[Pavel Machek]

Hi!

I tried cpufreq on hp omnibook with athlon4 cpu... Its interesting: It
works *in addition* to ACPI throttling. That means I can slow it down
to 4% of top speed.

Documentation/cpufreq should probably be renamed to Doc*/cpufreq.txt.

Kernel says:

powernow: AMD K7 CPU detected.
powernow: PowerNOW! Technology present. Can scale: frequency and
voltage.
powernow: Found PSB header at c00f7ab0
powernow: Table version: 0x12
powernow: Flags: 0x0 (Mobile voltage regulator)
powernow: Settling Time: 100 microseconds.
powernow: Has 14 PST tables. (Only dumping ones relevant to this CPU).
powernow: PST:2 (@c00f7ae4)
powernow:  cpuid: 0x761 fsb: 100        maxFID: 0xc     startvid: 0xc
powernow:    FID: 0x10 (3.0x [300MHz])  VID: 0xc (1.400V)
powernow:    FID: 0x4 (5.0x [500MHz])   VID: 0xc (1.400V)
powernow:    FID: 0x6 (6.0x [600MHz])   VID: 0xc (1.400V)
powernow:    FID: 0x8 (7.0x [700MHz])   VID: 0xc (1.400V)
powernow:    FID: 0xc (9.0x [900MHz])   VID: 0xc (1.400V)

First it claims it can scale voltage, then I see I can only use
1.4V. Too bad for me (and my batteries ;-)...
								Pavel
-- 
When do you have a heart between your knees?
[Johanka's followup: and *two* hearts?]

Re: cpufreq on athlon4

[Dave Jones]

On Sat, Feb 15, 2003 at 11:12:37PM +0100, Pavel Machek wrote:

 > Documentation/cpufreq should probably be renamed to Doc*/cpufreq.txt.

agreed.
 
 > powernow:  cpuid: 0x761 fsb: 100        maxFID: 0xc     startvid: 0xc
 > powernow:    FID: 0x10 (3.0x [300MHz])  VID: 0xc (1.400V)
 > powernow:    FID: 0x4 (5.0x [500MHz])   VID: 0xc (1.400V)
 > powernow:    FID: 0x6 (6.0x [600MHz])   VID: 0xc (1.400V)
 > powernow:    FID: 0x8 (7.0x [700MHz])   VID: 0xc (1.400V)
 > powernow:    FID: 0xc (9.0x [900MHz])   VID: 0xc (1.400V)
 > 
 > First it claims it can scale voltage, then I see I can only use
 > 1.4V. Too bad for me (and my batteries ;-)...

Some laptops have *really* crap PST tables. For the majority, they are
quite sane.  I'm collecting model names/numbers to feed back to AMD
so they can go beat up vendors.

Its likely at some point I'll implement a way to override using
the BIOS table too.

		Dave

-- 
| Dave Jones.        http://www.codemonkey.org.uk
| SuSE Labs

Re: cpufreq on athlon4

[Pavel Machek]

Hi!

>  > Documentation/cpufreq should probably be renamed to Doc*/cpufreq.txt.
> 
> agreed.
>  
>  > powernow:  cpuid: 0x761 fsb: 100        maxFID: 0xc     startvid: 0xc
>  > powernow:    FID: 0x10 (3.0x [300MHz])  VID: 0xc (1.400V)
>  > powernow:    FID: 0x4 (5.0x [500MHz])   VID: 0xc (1.400V)
>  > powernow:    FID: 0x6 (6.0x [600MHz])   VID: 0xc (1.400V)
>  > powernow:    FID: 0x8 (7.0x [700MHz])   VID: 0xc (1.400V)
>  > powernow:    FID: 0xc (9.0x [900MHz])   VID: 0xc (1.400V)
>  > 
>  > First it claims it can scale voltage, then I see I can only use
>  > 1.4V. Too bad for me (and my batteries ;-)...
> 
> Some laptops have *really* crap PST tables. For the majority, they are
> quite sane.  I'm collecting model names/numbers to feed back to AMD
> so they can go beat up vendors.

Its HP omnibook xe3.

> Its likely at some point I'll implement a way to override using
> the BIOS table too.

So you think that in fact I should be able to run at lower
voltage?

								Pavel
-- 
When do you have a heart between your knees?
[Johanka's followup: and *two* hearts?]

Re: cpufreq on athlon4

[Dave Jones]

On Sun, Feb 16, 2003 at 05:03:06PM +0100, Pavel Machek wrote:

 > > Its likely at some point I'll implement a way to override using
 > > the BIOS table too.
 > 
 > So you think that in fact I should be able to run at lower
 > voltage?

I've seen two different laptops from differnent vendors both with the
same CPU, with wildly differing PST tables, so its very likely
yes. And the fact that the powernow registers state that it
supports voltage scaling, and then not offering any in the PST
seems a bit silly.

		Dave

-- 
| Dave Jones.        http://www.codemonkey.org.uk