On Tue, Aug 6, 2019 at 6:39 AM Viresh Kumar <viresh.kumar@xxxxxxxxxx> wrote:
>
> On 05-08-19, 11:42, Rafael J. Wysocki wrote:
> > On Tuesday, July 23, 2019 8:14:05 AM CEST Viresh Kumar wrote:
> > > @@ -310,8 +339,11 @@ static int __init acpi_processor_driver_init(void)
> > > cpuhp_setup_state_nocalls(CPUHP_ACPI_CPUDRV_DEAD, "acpi/cpu-drv:dead",
> > > NULL, acpi_soft_cpu_dead);
> > >
> > > - acpi_thermal_cpufreq_init();
> > > - acpi_processor_ppc_init();
> > > + if (!cpufreq_register_notifier(&acpi_processor_notifier_block,
> > > + CPUFREQ_POLICY_NOTIFIER)) {
> > > + acpi_processor_cpufreq_init = true;
> >
> > Can't that be set/cleared by acpi_processor_notifier() itself?
>
> This is required to be done only once at initialization and setting it
> to true again and again on every invocation of the notifier callback
> doesn't look right.
>
> I have updated the patch based on rest of your suggestions, please see
> if it looks okay now.
Yes, it does, thanks!
[No need to resend, I'll take it from this message.]
