Re: Scaling governors in 14.1 stock kernel

On (13/11/13 14:16), Petri Kaukasoina <[email protected]> put forth the proposition:
>On Wed, Nov 13, 2013 at 10:12:35AM +0000, David Woodfall wrote:
>> I noticed that both my server and laptop have powersave as scaling
>> governor, and the choices available are that and performance.
>>
>> Can you guys take a look at the output of
>>
>> cat /sys/devices/system/cpu/cpu0/cpufreq/scaling_available_governors
>>
>> please?
>
>Does dmesg have lines like "Intel pstate controlling: cpu 0"? The
>config file has "CONFIG_X86_INTEL_PSTATE=y", so it doesn't use the
>traditional cpufreq drivers at all. Look at the three files under
>/sys/devices/system/cpu/intel_pstate
>You can control the intel_pstate driver by echoing values there.

Yes:

% dmesg | grep pstate
[    4.189351] Intel pstate controlling: cpu 0
[    4.192178] Intel pstate controlling: cpu 1


>      max_perf_pct: limits the maximum P state that will be requested by
>      the driver stated as a percentage of the avail performance.
>
>      min_perf_pct: limits the minimum P state that will be requested by
>      the driver stated as a percentage of the avail performance.
>
>      no_turbo: limits the driver to selecting P states below the turbo
>      frequency range.

Mine are 100, 57, 0

>If you echo performance to scaling_governor, then the three files
>above contain 100, 100, and 0. The default (powersave) has a lower
>minimum P state, so it can lower the frequency.

So powersave is the default now?