Re: [PATCH v6 5/5] ACPI : enable ACPI_PROCESSOR_IDLE on ARM64

[Sudeep Holla]

Hi Daniel,

On 27/06/16 15:33, Daniel Lezcano wrote:
> On 06/14/2016 04:48 PM, Sudeep Holla wrote:
> > Now that ACPI processor idle driver supports LPI(Low Power Idle), lets
> > enable ACPI_PROCESSOR_IDLE for ARM64 too.
> >
> > This patch just removes the IA64 and X86 dependency on
> > ACPI_PROCESSOR_IDLE
> >
> > Cc: linux-arm-kernel@xxxxxxxxxxxxxxxxxxx
> > Cc: "Rafael J. Wysocki" <rjw@xxxxxxxxxxxxx>
> > Signed-off-by: Sudeep Holla <sudeep.holla@xxxxxxx>
> > ---
>
> Hi Sudeep,
>
> now that ACPI processor supports ARM64 did you check the
> CPUIDLE_DRIVER_STATE_START trick in the code and its derivative ?

No, that is used only for C-State and ARM64 doesn't support it.
Patch 1/5 puts all the C-State code under #ifdef so that it's not
compiled on ARM64.

> I deleted the patch 2/5 but there is a place where:

Sorry, I don't follow what you mean by that.

> if (max_cstate=0)
>      max_cstate=1;
>
> Probably this is because the POLL state is inserted, so there is always
> an idle state. But for ARM, that is not the case.

Yes

> Also, there are some places where the idle state index begins to 1. I
> think it should be 0 for ARM.

Yes for LPI, it does start from 0.

--
Regards,
Sudeep