Re: I think there's an issue with e3f1164fc9e ("PM: EM: Support late CPUs booting and capacity adjustment") if there's "holes" in your CPU topology

[Kenneth Crudup <kenny@panix.com>]

I'll test this today and get back to you.

-K

On 9/4/25 10:40, Rafael J. Wysocki wrote:
> On Wed, Sep 3, 2025 at 8:53 PM Rafael J. Wysocki <rafael@kernel.org> wrote:
>>
>> On Wed, Sep 3, 2025 at 8:43 PM Kenneth Crudup <kenny@panix.com> wrote:
>>>
>>>
>>> Is there a way to distinguish between "offline cores" and
>>> "'non-'existent" cores?
>>>
>>> This way we could just skip the ones that "aren't" there, right?
>>
>> I'm not quite sure about the underlying use case TBH.
>>
>> The em_check_capacity_update() call may not be necessary on x86 at
>> all, but I need to double check.
> 
> So AFAICS, this can be addressed by something like the attached patch
> (the majority of changes in it is just moving kerneldoc comments
> around and function renaming).
> 
> Since intel_pstate handles capacity updates itself, it can do without
> em_check_capacity_update().

-- 
Kenneth R. Crudup / Sr. SW Engineer, Scott County Consulting, Orange 
County CA