Re: [PATCH 04/18] cpufreq: Manage fallback policies in a list

On Tuesday, February 17, 2015 01:36:47 PM Viresh Kumar wrote:

On 5 February 2015 at 23:11, Rafael J. Wysocki rjw@rjwysocki.net wrote:

...

I'm still concerned about the case when the last policy CPU is physically going away, in which we do the offline as a preliminary step and then will go for full CPU device unregistration.

I need more clarity on how this happens. So, on a physical hotplug out:

• We first do cpu offline, i.e. hotplug notifier gets called.

Yes.

• And then during unregistration the sysfs remove will get called?

The CPU device goes away then and that should trigger the remove of all the sysfs directories under it.