I haven't been able to set the chosen policy for CPU0

I haven't been able to set the chosen policy for CPU0

[Christian Flamm]

"I haven't been able to set the chosen policy for CPU0._I set 
1667000-1000000-ondemand_System says 1000000-1000000-ondemand_"

That's what the log files says. CPU's frequenzy randomly falls to minimum 
(1000000). Sometimes I can manually correct that, but more often I can't and 
the above message appears.

$ zcat /proc/config.gz | grep -i freq
	# CPU Frequency scaling
	CONFIG_CPU_FREQ=y
	CONFIG_CPU_FREQ_TABLE=y
	# CONFIG_CPU_FREQ_DEBUG is not set
	CONFIG_CPU_FREQ_STAT=y
	# CONFIG_CPU_FREQ_STAT_DETAILS is not set
	CONFIG_CPU_FREQ_DEFAULT_GOV_PERFORMANCE=y
	# CONFIG_CPU_FREQ_DEFAULT_GOV_USERSPACE is not set
	# CONFIG_CPU_FREQ_DEFAULT_GOV_ONDEMAND is not set
	# CONFIG_CPU_FREQ_DEFAULT_GOV_CONSERVATIVE is not set
	CONFIG_CPU_FREQ_GOV_PERFORMANCE=y
	CONFIG_CPU_FREQ_GOV_POWERSAVE=y
	CONFIG_CPU_FREQ_GOV_USERSPACE=y
	CONFIG_CPU_FREQ_GOV_ONDEMAND=y
	CONFIG_CPU_FREQ_GOV_CONSERVATIVE=y
	# CPUFreq processor drivers
	CONFIG_X86_ACPI_CPUFREQ=m
	# CONFIG_X86_CPUFREQ_NFORCE2 is not set
	# CONFIG_X86_ACPI_CPUFREQ_PROC_INTF is not set

$ cat /etc/cpufreqd.conf
	[General]
	pidfile=/var/run/cpufreqd.pid
	poll_interval=2
	verbosity=7
	double_check=1
	#enable_plugins=apci_ac,cpu,acpi_battery
	enable_remote=1
	remote_group=wheel
	[/General]
	
	[Profile]
	name=Halbgas
	minfreq=1000000
	maxfreq=1667000
	policy=ondemand
	[/Profile]
	
	[Profile]
	name=Vollgas
	minfreq=1667000
	maxfreq=1667000
	policy=performance
	[/Profile]
	
	[Rule]
	name=none
	ac=on
	profile=Vollgas
	[/Rule]

	[Rule]
	name=battery
	ac=off
	profile=Halbgas
	[/Rule]

Already upgraded to kernel 2.6.25-gentoo-r2 and cpufreqd 2.2.1... nothing 
changes. Processor is an Intel(R) Core(TM) Duo CPU L2400 @ 1.66GHz. Would you 
say thats probably a broken kernel or a broken bios?

Regards,
alcCapone.

I haven't been able to set the chosen policy for CPU0

[Christian Flamm]

I think I can now reproduce that misbehaviour. I happens when both of my cores 
are at 100%  - than min- *and* max-frequency fall down to minimum (1000000Hz)

$ cpufreq-info
	cpufrequtils 002: cpufreq-info (C) Dominik Brodowski 2004-2006
	Report errors and bugs to linux@brodo.de, please.
	analyzing CPU 0:
	  driver: acpi-cpufreq
	  CPUs which need to switch frequency at the same time: 0 1
	  hardware limits: 1000 MHz - 1.67 GHz
	  available frequency steps: 1.67 GHz, 1.33 GHz, 1000 MHz
	  available cpufreq governors: conservative, ondemand, userspace, powersave, 
performance
	  current policy: frequency should be within 1000 MHz and 1000 MHz.
	                  The governor "performance" may decide which speed to use
	                  within this range.
	  current CPU frequency is 1000 MHz (asserted by call to hardware).
	analyzing CPU 1:
	  driver: acpi-cpufreq
	  CPUs which need to switch frequency at the same time: 0 1
	  hardware limits: 1000 MHz - 1.67 GHz
	  available frequency steps: 1.67 GHz, 1.33 GHz, 1000 MHz
	  available cpufreq governors: conservative, ondemand, userspace, powersave, 	
performance	
	  current policy: frequency should be within 1000 MHz and 1000 MHz.
                  	The governor "performance" may decide which speed to use
        	        within this range.
	  current CPU frequency is 1000 MHz (asserted by call to hardware).

If I then try to force another policy manually by

(just to show you) 
$ cpufreqd-get
	Name (#1):      Halbgas
	Governor:       ondemand
	Min freq:       1000000
	Max freq:       1667000
	
	Name (#2):      Vollgas
	Active on CPU#: 0, 1
	Governor:       performance
	Min freq:       1667000
	Max freq:       1667000

$ cpufreqd-set 2

the log says: 
	"cpufreqd_set_profile     : Couldn't set profile "Vollgas" set for cpu 
(1667000-1667000-performance)_"

or

$ cpufreqd-set 1

the log says: 
	"I haven't been able to set the chosen policy for CPU0._I set 
1667000-1000000-ondemand_System says 1000000-1000000-ondemand_"

After a certain while (10 minutes) the manual commands work and stay for about 
5 minutes if both of my cores are fully stressed and than the whole story 
starts all over.

Regards, Christian.

RE: I haven't been able to set the chosen policy for CPU0

[Pallipadi, Venkatesh]

 

>-----Original Message-----
>From: cpufreq-bounces@lists.linux.org.uk 
>[mailto:cpufreq-bounces@lists.linux.org.uk] On Behalf Of 
>Christian Flamm
>Sent: Monday, May 19, 2008 9:42 AM
>To: cpufreq@lists.linux.org.uk
>Subject: I haven't been able to set the chosen policy for CPU0
>
>I think I can now reproduce that misbehaviour. I happens when 
>both of my cores 
>are at 100%  - than min- *and* max-frequency fall down to 
>minimum (1000000Hz)
>

Looks like the CPUs are being driver to lower freq due to temperature
limits. CPUs being busy is increasing the temp and they go to lower freq
until they reach the lower temp. Looking at
/proc/acpi/thermal_zone/*/*
when this happens can confirm this.

Thanks,
Venki

Re: I haven't been able to set the chosen policy for CPU0

[Christian Flamm]

>>I think I can now reproduce that misbehaviour. I happens when
>>both of my cores
>>are at 100%  - than min- *and* max-frequency fall down to
>>minimum (1000000Hz)

> Looks like the CPUs are being driver to lower freq due to temperature
> limits. CPUs being busy is increasing the temp and they go to lower freq
> until they reach the lower temp. Looking at
> /proc/acpi/thermal_zone/*/*
> when this happens can confirm this.

First of all, thx for the reply. 

Before, at, and after the freqency falls doen to minimum:

$ cat /proc/acpi/thermal_zone/THM0/state
state:                   ok

So its something else.

Regards, Christian.

Re: I haven't been able to set the chosen policy for CPU0

[Christian Flamm]

> >>>I think I can now reproduce that misbehaviour. I happens when
> >>>both of my cores
> >>>are at 100%  - than min- *and* max-frequency fall down to
> >>>minimum (1000000Hz)

> >> Looks like the CPUs are being driver to lower freq due to temperature
> >> limits. CPUs being busy is increasing the temp and they go

> >to lower freq

> >> until they reach the lower temp. Looking at
> >> /proc/acpi/thermal_zone/*/*
> >> when this happens can confirm this.

> >First of all, thx for the reply.

> >Before, at, and after the freqency falls doen to minimum:

> >$ cat /proc/acpi/thermal_zone/THM0/state
> >state:                   ok

> >So its something else.

> BTW, what does other values in THM0/* look like. Temp, passive, active,
> etc..

Maybe you're right.

Temperature increases up to 83°C and then the cpu seems to get auto-throttled.

$ ls /proc/acpi/thermal_zone/THM0/
cooling_mode  polling_frequency  state  temperature  trip_points

$cat /proc/acpi/thermal_zone/THM0/cooling_mode
<setting not supported>