Re: [PATCH 3/4] clk: scpi: add support for cpufreq virtual device

[Sudeep Holla]

On 21/05/15 00:45, Stephen Boyd wrote:
> On 04/27, Sudeep Holla wrote:
> > The clocks for the CPUs are provided by SCP and are managed by this
> > clock driver. So the cpufreq device needs to be added only after the
> > clock get registered and removed when this driver is unloaded.
> >
> > This patch manages the cpufreq virtual device based on the clock
> > availability.
> >
> > Signed-off-by: Sudeep Holla <sudeep.holla@xxxxxxx>
>
> The cpufreq device can't handle probe defer? I suppose we do it
> this way because we need to create a platform device somewhere
> and this is the best place to do so?

Correct, unless the communication with the SCP firmware is established
and the CPU clocks are registered, it makes no sense to register CPUFreq
virtual device. This seems to be best place IMO.

And yes, thanks to the absence of platform code on ARM64, where such
code was traditionally stashed away and got away with such scenarios.
Let me if you disagree with this or have any other suggestions.

Regards,
Sudeep
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/