Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1370012 > unrolled thread

Re: [intel-pstate driver regression] processor frequency very high even if in idle

Back to article view | Back to linux.kernel

This discussion starts older than the indexed window; earlier articles aren't shown. The article labeled Started by below is the oldest one visible, not the original post.

Contents

  Re: [intel-pstate driver regression] processor frequency very high  even if in idle Jörg Otte <jrg.otte@gmail.com> - 2016-04-02 19:20 +0200
    Re: [intel-pstate driver regression] processor frequency very high  even if in idle Sedat Dilek <sedat.dilek@gmail.com> - 2016-04-02 20:30 +0200
      RE: [intel-pstate driver regression] processor frequency very high even if in idle "Doug Smythies" <dsmythies@telus.net> - 2016-04-03 21:10 +0200
        Re: [intel-pstate driver regression] processor frequency very high  even if in idle Sedat Dilek <sedat.dilek@gmail.com> - 2016-04-04 07:20 +0200
          RE: [intel-pstate driver regression] processor frequency very high even if in idle "Doug Smythies" <dsmythies@telus.net> - 2016-04-04 08:20 +0200
            Re: [intel-pstate driver regression] processor frequency very high  even if in idle Sedat Dilek <sedat.dilek@gmail.com> - 2016-04-04 16:20 +0200

#1370012 — Re: [intel-pstate driver regression] processor frequency very high even if in idle

2016-04-02 17:28 GMT+02:00 Srinivas Pandruvada
<srinivas.pandruvada@linux.intel.com>:
>
> On Sat, 2016-04-02 at 08:30 +0200, Sedat Dilek wrote:
>> > I am trying CPU_FREQ_DEFAULT_GOV_PERFORMANCE=y from
>> > linux-pm.git#linux-next out of curiosity...
>> >
>> > $ ./scripts/diffconfig /boot/config-$(uname -r) .config
>> >  CPU_FREQ_DEFAULT_GOV_PERFORMANCE y -> n
>> > +CPU_FREQ_DEFAULT_GOV_SCHEDUTIL y
>> > +CPU_FREQ_GOV_ATTR_SET y
>> > +CPU_FREQ_GOV_SCHEDUTIL y
>> >
>> > ...will report.
>> >
>>
>> Not sure why I see here "powersave".
>> Does Intel-PState driver not support CPU_FREQ_GOV_SCHEDUTIL?
>>
>> $ cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_driver
>> intel_pstate
>> intel_pstate
>> intel_pstate
>> intel_pstate
>>
>> $ cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor
>> powersave
>> powersave
>> powersave
>> powersave
>
> If you are using Ubuntu, the OS has a script which will automatically
> change from performance.
> Doug can give more information on this script.
>
> Thanks,
> Srinivas
>
>
>
>>
>> See also attached files.
>>
>> - sed@ -

maybe:
/etc/init.d/ondemand

Thanks, Jörg

[toc] | [next] | [standalone]

#1370029

On Sat, Apr 2, 2016 at 7:19 PM, Jörg Otte <jrg.otte@gmail.com> wrote:
> 2016-04-02 17:28 GMT+02:00 Srinivas Pandruvada
> <srinivas.pandruvada@linux.intel.com>:
>>
>> On Sat, 2016-04-02 at 08:30 +0200, Sedat Dilek wrote:
>>> > I am trying CPU_FREQ_DEFAULT_GOV_PERFORMANCE=y from
>>> > linux-pm.git#linux-next out of curiosity...
>>> >
>>> > $ ./scripts/diffconfig /boot/config-$(uname -r) .config
>>> >  CPU_FREQ_DEFAULT_GOV_PERFORMANCE y -> n
>>> > +CPU_FREQ_DEFAULT_GOV_SCHEDUTIL y
>>> > +CPU_FREQ_GOV_ATTR_SET y
>>> > +CPU_FREQ_GOV_SCHEDUTIL y
>>> >
>>> > ...will report.
>>> >
>>>
>>> Not sure why I see here "powersave".
>>> Does Intel-PState driver not support CPU_FREQ_GOV_SCHEDUTIL?
>>>
>>> $ cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_driver
>>> intel_pstate
>>> intel_pstate
>>> intel_pstate
>>> intel_pstate
>>>
>>> $ cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor
>>> powersave
>>> powersave
>>> powersave
>>> powersave
>>
>> If you are using Ubuntu, the OS has a script which will automatically
>> change from performance.
>> Doug can give more information on this script.
>>
>> Thanks,
>> Srinivas
>>
>>
>>
>>>
>>> See also attached files.
>>>
>>> - sed@ -
>
> maybe:
> /etc/init.d/ondemand
>

With CONFIG_CPU_FREQ_DEFAULT_GOV_SCHEDUTIL=y (linux-pm.git#linux-next) I get...

$ cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_driver
acpi-cpufreq
acpi-cpufreq
acpi-cpufreq
acpi-cpufreq

$ cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor
ondemand
ondemand
ondemand
ondemand

...is there a difference when using intel_pstate as scaling_driver?
Are the scripts of Ubuntu working properly with acpi-cpufreq (only)?

- Sedat -

[toc] | [prev] | [next] | [standalone]

#1370274 — RE: [intel-pstate driver regression] processor frequency very high even if in idle

On 2016.04.02 11:21 Sedat Dilek wrote:
> On Sat, Apr 2, 2016 at 7:19 PM, Jörg Otte wrote:
>> 2016-04-02 17:28 GMT+02:00 Srinivas Pandruvada wrote:
>>>
>>> If you are using Ubuntu, the OS has a script which will automatically
>>> change from performance.
>>> Doug can give more information on this script.
>
>> maybe:
>> /etc/init.d/ondemand

Yes.

> With CONFIG_CPU_FREQ_DEFAULT_GOV_SCHEDUTIL=y (linux-pm.git#linux-next) I get...
>
> $ cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_driver
> acpi-cpufreq
> acpi-cpufreq
> acpi-cpufreq
> acpi-cpufreq
>
> $ cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor
> ondemand
> ondemand
> ondemand
> ondemand

Yes, those are the expected results for the acpi-cpufreq CPU frequency scaling driver.
You should be able to observe the governor set to sched util for the first minute
after re-boot and/or if you set it yourself after the /etc/init.d/ondemand script
has finished (i.e. more than 1 minute after re-boot.)

> ...is there a difference when using intel_pstate as scaling_driver?

Yes, but only because there are different available governors for the two drivers.

> Are the scripts of Ubuntu working properly with acpi-cpufreq (only)?

As far as I know the /etc/init.d/ondemand is working properly. It sets the acpi-cpufreq
driver to use the "ondemand" governor and it sets the intel_pstate driver to use the
"powersave" governor.

... Doug

[toc] | [prev] | [next] | [standalone]

#1370356

On Sun, Apr 3, 2016 at 8:59 PM, Doug Smythies <dsmythies@telus.net> wrote:
> On 2016.04.02 11:21 Sedat Dilek wrote:
>> On Sat, Apr 2, 2016 at 7:19 PM, Jörg Otte wrote:
>>> 2016-04-02 17:28 GMT+02:00 Srinivas Pandruvada wrote:
>>>>

Hi Doug,

are you involved in the Ubuntu-OS? Developer for Canonical?

>>>> If you are using Ubuntu, the OS has a script which will automatically
>>>> change from performance.
>>>> Doug can give more information on this script.
>>
>>> maybe:
>>> /etc/init.d/ondemand
>
> Yes.
>
>> With CONFIG_CPU_FREQ_DEFAULT_GOV_SCHEDUTIL=y (linux-pm.git#linux-next) I get...
>>
>> $ cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_driver
>> acpi-cpufreq
>> acpi-cpufreq
>> acpi-cpufreq
>> acpi-cpufreq
>>
>> $ cat /sys/devices/system/cpu/cpu*/cpufreq/scaling_governor
>> ondemand
>> ondemand
>> ondemand
>> ondemand
>
> Yes, those are the expected results for the acpi-cpufreq CPU frequency scaling driver.
> You should be able to observe the governor set to sched util for the first minute
> after re-boot and/or if you set it yourself after the /etc/init.d/ondemand script
> has finished (i.e. more than 1 minute after re-boot.)
>

I have hardcoded to use "schedutil" driver after one minute in
/etc/init.d/ondemand for testing-purposes and not to fall back to
"ondemand".

>> ...is there a difference when using intel_pstate as scaling_driver?
>
> Yes, but only because there are different available governors for the two drivers.
>
>> Are the scripts of Ubuntu working properly with acpi-cpufreq (only)?
>
> As far as I know the /etc/init.d/ondemand is working properly. It sets the acpi-cpufreq
> driver to use the "ondemand" governor and it sets the intel_pstate driver to use the
> "powersave" governor.
>

I haven't looked at this exactly.
The ondemand-script is not saying to fall back to "powersave" in case
of intel_pstate-driver (here on Ubuntu/precise).

- Sedat -

[toc] | [prev] | [next] | [standalone]

#1370378 — RE: [intel-pstate driver regression] processor frequency very high even if in idle

On 2016.03.04 22:14 Sedat Dilek wrote:
> On Sun, Apr 3, 2016 at 8:59 PM, Doug Smythies <dsmythies@telus.net> wrote:
>> On 2016.04.02 11:21 Sedat Dilek wrote:
>>> On Sat, Apr 2, 2016 at 7:19 PM, Jörg Otte wrote:
>>>> 2016-04-02 17:28 GMT+02:00 Srinivas Pandruvada wrote:
>>>>

> are you involved in the Ubuntu-OS? Developer for Canonical?

I try to help with Ubuntu as a community volunteer.
I am not a developer with Canonical.

... [cut]...

>> As far as I know the /etc/init.d/ondemand is working properly. It sets the acpi-cpufreq
>> driver to use the "ondemand" governor and it sets the intel_pstate driver to use the
>> "powersave" governor.
>>

> I haven't looked at this exactly.
> The ondemand-script is not saying to fall back to "powersave" in case
> of intel_pstate-driver (here on Ubuntu/precise).

It is likely that newer versions of /etc/init.d/ondemand script
were never backported to the older 12.04 precise release.
Why not? Well, because, as far as I know, that release never used the
intel_pstate driver by default and so never had to fall through
to the powersave alternative.

Earlier on this thread, Srinivas correctly asked Jörg what distribution
was being used, so as to hopefully make it easier to reproduce the issue.
If we want to continue this distro specific conversation, perhaps
we should move it off-list or to some Ubuntu specific forum.

... Doug

[toc] | [prev] | [next] | [standalone]

#1370658

On Mon, Apr 4, 2016 at 8:14 AM, Doug Smythies <dsmythies@telus.net> wrote:
> On 2016.03.04 22:14 Sedat Dilek wrote:
>> On Sun, Apr 3, 2016 at 8:59 PM, Doug Smythies <dsmythies@telus.net> wrote:
>>> On 2016.04.02 11:21 Sedat Dilek wrote:
>>>> On Sat, Apr 2, 2016 at 7:19 PM, Jörg Otte wrote:
>>>>> 2016-04-02 17:28 GMT+02:00 Srinivas Pandruvada wrote:
>>>>>
>
>> are you involved in the Ubuntu-OS? Developer for Canonical?
>
> I try to help with Ubuntu as a community volunteer.
> I am not a developer with Canonical.
>
> ... [cut]...
>
>>> As far as I know the /etc/init.d/ondemand is working properly. It sets the acpi-cpufreq
>>> driver to use the "ondemand" governor and it sets the intel_pstate driver to use the
>>> "powersave" governor.
>>>
>
>> I haven't looked at this exactly.
>> The ondemand-script is not saying to fall back to "powersave" in case
>> of intel_pstate-driver (here on Ubuntu/precise).
>
> It is likely that newer versions of /etc/init.d/ondemand script
> were never backported to the older 12.04 precise release.
> Why not? Well, because, as far as I know, that release never used the
> intel_pstate driver by default and so never had to fall through
> to the powersave alternative.
>
> Earlier on this thread, Srinivas correctly asked Jörg what distribution
> was being used, so as to hopefully make it easier to reproduce the issue.
> If we want to continue this distro specific conversation, perhaps
> we should move it off-list or to some Ubuntu specific forum.
>

Yes, it is getting off-topic.
Sorry for this.
I know I am here on my own and I think how to fix this.
( Honestly, I forgot about the initial issue. )

- Sedat -

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web