Re: [PATCH 1/1] cpufreq: Set policy->min and max as real QoS constraints

[Pierre Gondois <pierre.gondois@arm.com>]

Hello Sumit,

On 4/28/26 18:37, Sumit Gupta wrote:
>
> On 23/04/26 14:17, Pierre Gondois wrote:
>> External email: Use caution opening links or attachments
>>
>>
>> cpufreq_set_policy() will ultimately override the policy min/max
>> values written in the .init() callback through:
>> cpufreq_policy_online()
>> \-cpufreq_init_policy()
>>    \-cpufreq_set_policy()
>>      \-/* Set policy->min/max */
>> Thus the policy min/max values provided are only temporary.
>>
>> There is an exception if CPUFREQ_NEED_INITIAL_FREQ_CHECK is set and:
>> cpufreq_policy_online()
>> \-cpufreq_init_policy()
>>    \-__cpufreq_driver_target()
>>      \-cpufreq_driver->target()
>> is called. To avoid any regression, set policy->min/max in cpufreq.c
>> if the values were not initialized.
>>
>> In this patch:
>> - Setting policy->min or max value in driver .init() cb is
>>    interpreted as setting a QoS constraint.
>> - Remove policy->min/max initialization in drivers if the values
>>    are similar to policy->cpuinfo.min_freq/max_freq.
>>    The only drivers where these values are different are:
>>    - gx-suspmod.c
>>    - cppc-cpufreq.c
>>    - longrun.c
>> - For the cppc-cpufreq driver, the lowest non-linear freq. is
>>    used as a min QoS constraint as suggested at:
>> https://lore.kernel.org/lkml/20260213100633.15413-1-zhangpengjie2@huawei.com/
>>
>> Signed-off-by: Pierre Gondois <pierre.gondois@arm.com>
>
> Tested-by: Sumit Gupta <sumitg@nvidia.com>
> Reviewed-by: Sumit Gupta <sumitg@nvidia.com>
Thanks!
There might be a v2 with some small modifications.