Thinkpad T61 consuming more power on AC

Hello,
may I please again ask for your help regarding this issue:
I notice, with 2.6.25, that my T61 runs rather quiet on battery while in AC mode the fan runs far often, if not all the time.

With recent powertop I could find out the most likely reason:

(Kernel 2.6.25)

AC:
Your CPU supports the following C-states : C1 C2 C3 C4 C5 C6 Your BIOS reports the following C-states : C1 C2 C3

Battery:
Your CPU supports the following C-states : C1 C2 C3 C4 C5 C6 Your BIOS reports the following C-states : C1 C2 C6

Powertop then shows me it's actually using the processor states as listed as reported from BIOS.
So obviously it doesn't use the low power state C6 when on AC while it does on battery.
I already tried all reasonable as well as non reasonable BIOS settings to see if that changes something, but no, I can only get it worse in battery mode, but not any better in AC mode.

Is it possible, maybe with some slight tweaks (setpci...), to convice the kernel (or BIOS?) to also use that C6 state when in AC mode?

I would love to!

Thanks,
Daniel

--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

Understand what you said. In fact OS won't use the CPU C3/C4/C5/C6 directly. In different cases they will be mapped to ACPI C3 state , which means the low power state that is used by OS. For example: On the T61 laptop: In AC mode, CPU C3 is mapped to ACPI C3; In battery mode:
CPU C6 is mapped to ACPI C3.
The C-state mapping between CPU C3/C4/C5/C6 and ACPI C3 is defined
in ACPI tables, which is related with the BIOS. In order to use CPU C6 state in AC mode, we must modify the C-state mapping in ACPI tables. It is not easy. And IMO it is meaningless(CPU C6 can save more power than CPU C3. But it is not obvious). So not recommend to do so.
Of course if still hope to do so, please refer to the _CST object
defined in section 8 of ACPI 3.0 spec and custom DSDT should be used.

Thanks.
Yakui

--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

Am Dienstag, den 08.07.2008, 09:31 +0800 schrieb Zhao Yakui:

Thanks Yakui. I have to say that I don't have any ACPI knowledge other than common hearsay so I have to learn ACPI from the beginning and I hope I could find some spare time to get and idea, at least.

- Extract the ACPI tables of my notebook
- Read the spec and tweak the right entry
- Convince the kernel to use the customized table

Probably I could also read the ACPI tables when on battery just to load it when in AC mode? (I could then skip step 2 of understanding what I'm doing... ;-)

For the effectiveness:
I cannot say whether the power consumption is that big, the difference of the fan (not) getting on my nerves _is_ significant however. I just had the idea that the behaviour of the embedded controller for controlling the fan may also be different in AC vs. battery mode (the threshold temperatures), so I it's indeed not garantueed that using C6 vs. C3 alone does the trick. But I do hope so.

Thanks,
Daniel