* cpufreq - powernv driver
@ 2022-06-27 13:24 Anna-Maria Behnsen
2022-09-26 13:20 ` Anna-Maria Behnsen
0 siblings, 1 reply; 3+ messages in thread
From: Anna-Maria Behnsen @ 2022-06-27 13:24 UTC (permalink / raw)
To: linux-pm; +Cc: rafael, Thomas Gleixner, Viresh Kumar, Michael Ellerman
Hi,
during my work on a timer hierarchy I stumbled over powernv cpufreq driver
which is using a timer which is deferrable and pinned. This is the only
timer in kernel which uses this combination and I would like to get rid of
it.
Only removing the pinned or deferrable flag could not be the proper
solution, right?
I'm not familiar with cpufreq. I was wondering if it's possible to rework
the powernv cpufreq driver to use cpufreq infrastructure instead handling
it's own timer for ramping down? I would be happy if someone could help me
with some pointers how this could be solved.
Thanks,
Anna-Maria
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: cpufreq - powernv driver
2022-06-27 13:24 cpufreq - powernv driver Anna-Maria Behnsen
@ 2022-09-26 13:20 ` Anna-Maria Behnsen
2022-09-26 13:40 ` Rafael J. Wysocki
0 siblings, 1 reply; 3+ messages in thread
From: Anna-Maria Behnsen @ 2022-09-26 13:20 UTC (permalink / raw)
To: linux-pm; +Cc: rafael, Thomas Gleixner, Viresh Kumar, Michael Ellerman
ping?
On Mon, 27 Jun 2022, Anna-Maria Behnsen wrote:
> Hi,
>
> during my work on a timer hierarchy I stumbled over powernv cpufreq driver
> which is using a timer which is deferrable and pinned. This is the only
> timer in kernel which uses this combination and I would like to get rid of
> it.
>
> Only removing the pinned or deferrable flag could not be the proper
> solution, right?
>
> I'm not familiar with cpufreq. I was wondering if it's possible to rework
> the powernv cpufreq driver to use cpufreq infrastructure instead handling
> it's own timer for ramping down? I would be happy if someone could help me
> with some pointers how this could be solved.
>
> Thanks,
>
> Anna-Maria
>
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: cpufreq - powernv driver
2022-09-26 13:20 ` Anna-Maria Behnsen
@ 2022-09-26 13:40 ` Rafael J. Wysocki
0 siblings, 0 replies; 3+ messages in thread
From: Rafael J. Wysocki @ 2022-09-26 13:40 UTC (permalink / raw)
To: Anna-Maria Behnsen
Cc: Linux PM, Rafael J. Wysocki, Thomas Gleixner, Viresh Kumar,
Michael Ellerman
On Mon, Sep 26, 2022 at 3:20 PM Anna-Maria Behnsen
<anna-maria@linutronix.de> wrote:
>
> ping?
Sorry, this fell off my radar.
> On Mon, 27 Jun 2022, Anna-Maria Behnsen wrote:
>
> > Hi,
> >
> > during my work on a timer hierarchy I stumbled over powernv cpufreq driver
> > which is using a timer which is deferrable and pinned. This is the only
> > timer in kernel which uses this combination and I would like to get rid of
> > it.
> >
> > Only removing the pinned or deferrable flag could not be the proper
> > solution, right?
> >
> > I'm not familiar with cpufreq. I was wondering if it's possible to rework
> > the powernv cpufreq driver to use cpufreq infrastructure instead handling
> > it's own timer for ramping down? I would be happy if someone could help me
> > with some pointers how this could be solved.
cpufreq drivers need not use timers at all in general.
What they need to do is to register a callback via
cpufreq_add_update_util_hook(), like intel_pstate or the schedutil
governor.
I'll look at the powernv driver later this week.
> > Thanks,
> >
> > Anna-Maria
> >
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2022-09-26 15:08 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-06-27 13:24 cpufreq - powernv driver Anna-Maria Behnsen
2022-09-26 13:20 ` Anna-Maria Behnsen
2022-09-26 13:40 ` Rafael J. Wysocki
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).