Re: [PATCH v3 6/7] thermal/drivers/cpu_cooling: Introduce the cpu idle cooling driver

From: Sudeep Holla <sudeep.holla@arm.com>
To: Daniel Lezcano <daniel.lezcano@linaro.org>
Cc: Viresh Kumar <viresh.kumar@linaro.org>,
	edubezval@gmail.com, kevin.wangtao@linaro.org,
	leo.yan@linaro.org, vincent.guittot@linaro.org,
	linux-kernel@vger.kernel.org, javi.merino@kernel.org,
	rui.zhang@intel.com, daniel.thompson@linaro.org,
	linux-pm@vger.kernel.org,
	Amit Daniel Kachhap <amit.kachhap@gmail.com>
Subject: Re: [PATCH v3 6/7] thermal/drivers/cpu_cooling: Introduce the cpu idle cooling driver
Date: Mon, 16 Apr 2018 14:03:48 +0100	[thread overview]
Message-ID: <20180416130348.GA5120@e107533-lin> (raw)
In-Reply-To: <1d4ce2d7-cc24-83e1-dbf9-2c1cd2f5ca74@linaro.org>

On Mon, Apr 16, 2018 at 02:49:35PM +0200, Daniel Lezcano wrote:
> On 16/04/2018 14:31, Sudeep Holla wrote:
> > On Mon, Apr 16, 2018 at 02:10:30PM +0200, Daniel Lezcano wrote:
> >> On 16/04/2018 12:10, Viresh Kumar wrote:
> >>> On 16-04-18, 12:03, Daniel Lezcano wrote:
> >>>> On 16/04/2018 11:50, Viresh Kumar wrote:
> >>>>> On 16-04-18, 11:45, Daniel Lezcano wrote:
> >>>>>> Can you elaborate a bit ? I'm not sure to get the point.
> >>>>>
> >>>>> Sure. With your current code on Hikey960 (big/LITTLE), you end up
> >>>>> creating two cooling devices, one for the big cluster and one for
> >>>>> small cluster. Which is the right thing to do, as we also have two
> >>>>> cpufreq cooling devices.
> >>>>>
> >>>>> But with the change Sudeep is referring to, the helper you used to get
> >>>>> cluster id will return 0 (SoC id) for all the 8 CPUs. So your code
> >>>>> will end up creating a single cpuidle cooling device for all the CPUs.
> >>>>> Which would be wrong.
> >>>>
> >>>> Is the semantic of topology_physical_package_id changing ?
> >>>
> >>> That's what I understood from his email.
> >>>
> >>>> I don't
> >>>> understand the change Sudeep is referring to.
> >>
> >> Actually there is no impact with the change Sudeep is referring to. It
> >> is for ACPI, we are DT based. Confirmed with Jeremy.
> >>
> > 
> > No, it will change for DT. The aim is to be consistent irrespective of
> > h/w or f/w description(i.e ADCPI or DT)
> 
> What will happen with the code using topology_physical_package_id ?
> 
> drivers/acpi/processor_thermal.c:       int id = topology_physical_package_id(cpu);
> drivers/acpi/processor_thermal.c:               if (topology_physical_package_id(i) == id)
> drivers/acpi/processor_thermal.c:               if (topology_physical_package_id(i) ==
> drivers/acpi/processor_thermal.c: topology_physical_package_id(cpu))
> 

ACPI and hence will have right notion of package.

> drivers/block/mtip32xx/mtip32xx.c: topology_physical_package_id(cpumask_first(node_mask)),
> 

Not sure if it was ever used on ARM/ARM64

> drivers/cpufreq/arm_big_little.c:       return topology_physical_package_id(cpu);

Yes this is main affected driver. I don't think it's used by any ARM64
platform anymore. This is one of the reason I divorced SCPI driver from
this recently. ARM32 may still retain old definition of package_id(not
sure, depends if anyone as need to change that and maintainers view on
that). If not, we need to fix this driver along with the code changing
the definition.

> 
> drivers/crypto/qat/qat_common/adf_common_drv.h: return topology_physical_package_id(smp_processor_id());
> drivers/crypto/virtio/virtio_crypto_common.h:   node = topology_physical_package_id(cpu);
> 

x86 specific and never assumed cluster id.

> kernel/events/core.c:           event_pkg = topology_physical_package_id(event_cpu);
> kernel/events/core.c:           local_pkg = topology_physical_package_id(local_cpu);
> 

Generic and hence never assumed cluster id.

--
Regards,
Sudeep