8 cstates but only 4 to use?

8 cstates but only 4 to use?

[David W. Hankins]

I looked through the excellent Howto/FAQ provided by Glenn Ariel (thank
you, sir), and he answered a lot of my questions, but was unable to find
the answer to one remaining niggling curiosity.

The 'power' procfile lists a max_cstate of C8, but only lists C1-C4 as
usable states to enter...I notice it is also this way in the example
file output provided in Mr. Ariel's howto, so this must not be uncommon
or new.

But, if 8 states are avaialble, why are only 4 in use?

>From a cursory investigation of the sources, it seems through #def's,
this is a fixed upper limit.

And I figure it can't be just that no one ever thought to #define 5-9,
that would be silly.

Thanks in advance, sorry if this is in a FAQ elsewhere, google is getting
a lot of folks' dmesg output when I try to find answers along these lines.

-- 
David W. Hankins		"If you don't do it right the first time,
Software Engineer			you'll just have to do it again."
Internet Systems Consortium, Inc.		-- Jack T. Hankins

Re: 8 cstates but only 4 to use?

[Dominik Brodowski]

Hi,

On Sat, Mar 04, 2006 at 05:54:23AM +0000, David W. Hankins wrote:
> The 'power' procfile lists a max_cstate of C8, but only lists C1-C4 as
> usable states to enter...I notice it is also this way in the example
> file output provided in Mr. Ariel's howto, so this must not be uncommon
> or new.

max_cstate is the maximum the kernel would make use of if they were provided
by the hardware and by the BIOS. However,

> But, if 8 states are avaialble, why are only 4 in use?

it is quite common that only one, two, three or -- at most! -- four of these
states are actually available on a system. The ACPI core allows for eight to
be ready for future systems which might have more than these four available.

	Dominik