Re: [PATCH v14 5/5] cpufreq: Only disable boost during cpu online when using frequency tables

[Mario Limonciello <mario.limonciello@amd.com>]

On 6/25/2024 22:27, Viresh Kumar wrote:
> On 26-06-24, 08:55, Viresh Kumar wrote:
>> On 25-06-24, 22:20, Mario Limonciello wrote:
>>> Yeah; that's effectively the same result as Gautham's suggestion.  He had
>>> just said to change policy_has_boost_freq() for the same.
>>
>> I know. The problem is policy_has_boost_freq() (as its name suggests)
>> needs to check if the policy supports boost freqs (in a generic way)
>> and it is used at several other places and it would be wrong to hack
>> that routine to fix this problem.
>>
>> All we want here is for the core to not touch boost_enabled if the
>> policy->init() function has already done so.
>>
>>> I'll test it with yours and reconcile the better one to submit back out for
>>> v15.
>>
>> You can send it separately to be honest, and it looks like a fix, so
>> Rafael should be able to get it merged a bit sooner. Add a fixes tag
>> too.
> 
> And maybe send patch for acpi-cpufreq and any other platform that is
> broken too..
> 

I can take it all through the amd-pstate tree.
I'll put it at the front of the series.
I think as long as we can get it merged before ~rc6 it should be fine 
for the 6.11 merge window.