RE: [RFC] Documentation: Add documentation for new performance_profile sysfs class

[Barnabás Pőcze]

Hi


2020. október 5., hétfő 14:58 keltezéssel, Limonciello, Mario írta:
> > On modern systems CPU/GPU/... performance is often dynamically configurable
> > in the form of e.g. variable clock-speeds and TPD. The performance is often
> > automatically adjusted to the load by some automatic-mechanism (which may
> > very well live outside the kernel).
> > These auto performance-adjustment mechanisms often can be configured with
> > one of several performance-profiles, with either a bias towards low-power
> > consumption (and cool and quiet) or towards performance (and higher power
> > consumption and thermals).
> > Introduce a new performance_profile class/sysfs API which offers a generic
> > API for selecting the performance-profile of these automatic-mechanisms.
>
> If introducing an API for this - let me ask the question, why even let each
> driver offer a class interface and userspace need to change "each" driver's
> performance setting?
>
> I would think that you could just offer something kernel-wide like
> /sys/power/performance-profile
>
> Userspace can read and write to a single file. All drivers can get notified
> on this sysfs file changing.
>

That makes sense, in my opinion, from the regular user's perspective:
one switch to rule them all, no fuss. However, I don't think that scales well.
What if the hypothetical users wants to run a CPU-heavy workload, and thus wants
to put the GPU into "low-power" mode and the CPU into "performance" mode? What if
the users wants to put one GPU into "low-power" mode, but the other one into
"performance"? With the current specification, the user's needs could be easily
satisfied. I don't see how that's possible with a single switch. Nonetheless, I think
that a single global switch *in addition* to the class devices could possibly
simplify the userspace-kernel interaction for most users.


> The systems that react in firmware (such as the two that prompted
> this discussion) can change at that time. It leaves the possibility for a
> more open kernel implementation that can do the same thing though too by
> directly modifying device registers instead of ACPI devices.
>

Excuse my ignorance, but I don't really see why this interface would be tied to
ACPI devices? Why is it not possible to write a driver that implements this interface
and directly modifies device registers? Am I missing something obvious here?


> [...]


Thanks,
Barnabás Pőcze