Never in C3

Never in C3

[Fred Labrosse]

All,

The following seems to suggest that my processor is never in C3 state:

# cat /proc/acpi/processor/CPU0/power
active state:            C1
default state:           C1
bus master activity:     db66ffff
states:
   *C1:                  promotion[C2] demotion[--] latency[000] usage[02058670]
    C2:                  promotion[C3] demotion[C1] latency[020] usage[19743630]
    C3:                  promotion[--] demotion[C2] latency[200] usage[00000000]

Right?

I did read at some point that this was because of hotplug.  I thus stopped
hotplug and, at that time, I did see my processor in C3 state.  Hoever,
since then, I've never seen that again, and hotplug is not running.

And yes, C3 is claimed to be supported (from dmesg:

ACPI: Processor [CPU0] (supports C1 C2 C3, 8 throttling states)
)

Anything else that could cause that?  Any hints?

Cheers,

Fred



-------------------------------------------------------
This SF.Net email is sponsored by the new InstallShield X.
>From Windows to Linux, servers to mobile, InstallShield X is the
one installation-authoring solution that does it all. Learn more and
evaluate today! http://www.installshield.com/Dev2Dev/0504

Re: Never in C3

[Karol Kozimor]

Thus wrote Fred Labrosse:
> I did read at some point that this was because of hotplug.  I thus stopped
> hotplug and, at that time, I did see my processor in C3 state.  Hoever,
> since then, I've never seen that again, and hotplug is not running.

hotplug itself doesn't do any harm, though USB does. Possibly depending on
your host controller, you can't have anything plugged in and you might have
to reload the module (under 2.4) to have C3 working.
Best regards,

-- 
Karol 'sziwan' Kozimor
sziwan-DETuoxkZsSqrDJvtcaxF/A@public.gmane.org


-------------------------------------------------------
This SF.Net email is sponsored by the new InstallShield X.
>From Windows to Linux, servers to mobile, InstallShield X is the
one installation-authoring solution that does it all. Learn more and
evaluate today! http://www.installshield.com/Dev2Dev/0504

Re: Never in C3

[Fred Labrosse]

Karol Kozimor writes:
 > Thus wrote Fred Labrosse:
 > > I did read at some point that this was because of hotplug.  I thus stopped
 > > hotplug and, at that time, I did see my processor in C3 state.  Hoever,
 > > since then, I've never seen that again, and hotplug is not running.
 > 
 > hotplug itself doesn't do any harm, though USB does.

That is right.  At least this rings a bell.

 > Possibly depending on
 > your host controller, you can't have anything plugged in and you might have
 > to reload the module (under 2.4) to have C3 working.

Well, I have no usb device plugged in, no usb module loaded (I just made
sure that no usb bits were in the kernel) and still no C3.  What do you
mean by "you might have to reload the module"?  Which module.

BTW, I run vanilla 2.6.6. on an ACER TM636.

TIA,

Fred



-------------------------------------------------------
This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference
Learn from the experts at JavaOne(SM), Sun's Worldwide Java Developer
Conference, June 28 - July 1 at the Moscone Center in San Francisco, CA
REGISTER AND SAVE! http://java.sun.com/javaone/sf Priority Code NWMGYKND

Re: Never in C3

[Karol Kozimor]

Thus wrote Fred Labrosse:
>  > Possibly depending on
>  > your host controller, you can't have anything plugged in and you might have
>  > to reload the module (under 2.4) to have C3 working.
> Well, I have no usb device plugged in, no usb module loaded (I just made
> sure that no usb bits were in the kernel) and still no C3.  What do you
> mean by "you might have to reload the module"?  Which module.

The host controller module. Load it and see if it helps. Also, try running
a bare system (runlevel 1 or even init=/bin/bash) and add bits until you
hit something. Possible factors include sound (running a sound server? tell
it to suspend and let go of the /dev/* files), 3D, hard disk or network
activity...

> BTW, I run vanilla 2.6.6. on an ACER TM636.

Then you should be more or less fine, once you identify the specific
problem.
Best regards,

-- 
Karol 'sziwan' Kozimor
sziwan-DETuoxkZsSqrDJvtcaxF/A@public.gmane.org


-------------------------------------------------------
This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference
Learn from the experts at JavaOne(SM), Sun's Worldwide Java Developer
Conference, June 28 - July 1 at the Moscone Center in San Francisco, CA
REGISTER AND SAVE! http://java.sun.com/javaone/sf Priority Code NWMGYKND

Re: Never in C3

[Matthew Garrett]

On Mon, 2004-06-14 at 14:06 +0100, Fred Labrosse wrote:

>    *C1:                  promotion[C2] demotion[--] latency[000] usage[02058670]
>     C2:                  promotion[C3] demotion[C1] latency[020] usage[19743630]
>     C3:                  promotion[--] demotion[C2] latency[200] usage[00000000]

My recollection is that if the latency is larger than 100, Linux will
never use C3.

-- 
Matthew Garrett | mjg59-1xO5oi07KQx4cg9Nei1l7Q@public.gmane.org



-------------------------------------------------------
This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference
Learn from the experts at JavaOne(SM), Sun's Worldwide Java Developer
Conference, June 28 - July 1 at the Moscone Center in San Francisco, CA
REGISTER AND SAVE! http://java.sun.com/javaone/sf Priority Code NWMGYKND

Re: Never in C3

[Nate Lawson]

On Tue, 15 Jun 2004, Matthew Garrett wrote:
> On Mon, 2004-06-14 at 14:06 +0100, Fred Labrosse wrote:
>
> >    *C1:                  promotion[C2] demotion[--] latency[000] usage[02058670]
> >     C2:                  promotion[C3] demotion[C1] latency[020] usage[19743630]
> >     C3:                  promotion[--] demotion[C2] latency[200] usage[00000000]
>
> My recollection is that if the latency is larger than 100, Linux will
> never use C3.

The C2 limit is 100, C3 is 1000.

-Nate


-------------------------------------------------------
This SF.Net email is sponsored by The 2004 JavaOne(SM) Conference
Learn from the experts at JavaOne(SM), Sun's Worldwide Java Developer
Conference, June 28 - July 1 at the Moscone Center in San Francisco, CA
REGISTER AND SAVE! http://java.sun.com/javaone/sf Priority Code NWMGYKND

Re: Never in C3

[Frank Mehnert]

[-- Attachment #1: Type: text/plain, Size: 1294 bytes --]

Hi,

On Monday 14 June 2004 15:06, Fred Labrosse wrote:
> The following seems to suggest that my processor is never in C3 state:

BTW: It seems that my processor does not use the C1 state!

$ cat /proc/acpi/processor/CPU/power:
active state:            C2
default state:           C1
bus master activity:     ffffffff
states:
    C1:      promotion[C2] demotion[--] latency[000] usage[00000010]
                                                           ^^^^^^^^
   *C2:      promotion[C3] demotion[C1] latency[001] usage[04683642]
    C3:      promotion[--] demotion[C2] latency[085] usage[00004033]

Hardware: T40p with Pentium-M 1.6 GHz.
Everything seems to work perfectly well.

$ cat /proc/acpi/processor/CPU/performance:
state count:             6
active state:            P0
states:
   *P0:                  1600 MHz, 24500 mW, 500 uS
    P1:                  1400 MHz, 21000 mW, 500 uS
    P2:                  1200 MHz, 17000 mW, 500 uS
    P3:                  1000 MHz, 13500 mW, 500 uS
    P4:                  800 MHz, 9500 mW, 500 uS
    P5:                  600 MHz, 6000 mW, 500 uS

Frank
-- 
## Dept. of Computer Science, Dresden University of Technology, Germany ##
## http://os.inf.tu-dresden.de/~fm3                                     ##

[-- Attachment #2: signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]