On Wed, May 2, 2018 at 5:39 PM, Prakash, Prashanth
<pprakash@xxxxxxxxxxxxxx> wrote:
> Hi Viresh,
>
>
> On 4/27/2018 11:35 AM, Prashanth Prakash wrote:
>> Add support to specify platform specific transition_delay_us instead
>> of using the transition delay derived from PCC.
>>
>> With commit "3d41386d556d: cpufreq: CPPC: Use transition_delay_us
>> depending transition_latency" we are setting transition_delay_us
>> directly and not applying the LATENCY_MULTIPLIER. With this on Qualcomm
>> Centriq we can end up with a very high rate of frequency change requests
>> when using schedutil governor (default rate_limit_us=10 compared to an
>> earlier value of 10000).
>>
>> The PCC subspace describes the rate at which the platform can accept
>> commands on the CPPC's PCC channel. This includes read and write
>> command on the PCC channel that can be used for reasons other than
>> frequency transitions. Moreover the same PCC subspace can be used by
>> multiple freq domains and deriving transition_delay_us from it as we do
>> now can be sub-optimal.
>>
>> Moreover if a platform does not use PCC for desired_perf register then
>> there is no way to compute the transition latency or the delay_us.
>>
>> CPPC does not have a standard defined mechanism to get the transition
>> rate or the latency at the moment.
>>
>> Given the above limitations, it is simpler to have a platform specific
>> transition_delay_us and rely on PCC derived value only if a platform
>> specific value is not available.
>>
>> Signed-off-by: Prashanth Prakash <pprakash@xxxxxxxxxxxxxx>
>> Cc: Viresh Kumar <viresh.kumar@xxxxxxxxxx>
>> Cc: Rafael J. Wysocki <rjw@xxxxxxxxxxxxx>
>> Cc: 4.14+ <stable@xxxxxxxxxxxxxxx>
>> Fixes: 3d41386d556d ("cpufreq: CPPC: Use transition_delay_us depending
>> transition_latency)
>> ---
>> v2:
>> * Return final delay_us from cppc_cpufreq_get_transition_delay_us (Viresh)
>> v3 and v4:
>> * code style changes (Viresh)
>> ---
>> drivers/cpufreq/cppc_cpufreq.c | 47 ++++++++++++++++++++++++++++++++++++++++--
>> 1 file changed, 45 insertions(+), 2 deletions(-)
>
> Any inputs on this patch? I was hoping to get this in for 4.17 as it fixes undesirable(very low)
> transition delay when using schedutil governor on centriq platform.
I've queued it up for -rc4. I have a plan to push it later today.
Thanks,
Rafael
