Re: [PATCH v2] cpufreq: intel_pstate: Change the calculation of next pstate

2014-05-02 Thread Stratos Karafotis
On 02/05/2014 03:26 μμ, Rafael J. Wysocki wrote: > On Thursday, May 01, 2014 06:48:08 PM Dirk Brandewie wrote: >> On 05/01/2014 04:18 PM, Rafael J. Wysocki wrote: >>> On Thursday, May 01, 2014 02:30:42 PM Dirk Brandewie wrote: On 05/01/2014 02:00 PM, Stratos Karafotis wrote: > Currently th

Re: [PATCH v2] cpufreq: intel_pstate: Change the calculation of next pstate

2014-05-02 Thread Rafael J. Wysocki
On Thursday, May 01, 2014 06:48:08 PM Dirk Brandewie wrote: > On 05/01/2014 04:18 PM, Rafael J. Wysocki wrote: > > On Thursday, May 01, 2014 02:30:42 PM Dirk Brandewie wrote: > >> On 05/01/2014 02:00 PM, Stratos Karafotis wrote: > >>> Currently the driver calculates the next pstate proportional to

Re: [PATCH v2] cpufreq: intel_pstate: Change the calculation of next pstate

2014-05-01 Thread Dirk Brandewie
On 05/01/2014 04:18 PM, Rafael J. Wysocki wrote: On Thursday, May 01, 2014 02:30:42 PM Dirk Brandewie wrote: On 05/01/2014 02:00 PM, Stratos Karafotis wrote: Currently the driver calculates the next pstate proportional to core_busy factor, scaled by the ratio max_pstate / current_pstate. Using

Re: [PATCH v2] cpufreq: intel_pstate: Change the calculation of next pstate

2014-05-01 Thread Rafael J. Wysocki
On Thursday, May 01, 2014 02:30:42 PM Dirk Brandewie wrote: > On 05/01/2014 02:00 PM, Stratos Karafotis wrote: > > Currently the driver calculates the next pstate proportional to > > core_busy factor, scaled by the ratio max_pstate / current_pstate. > > > > Using the scaled load (core_busy) to calc

Re: [PATCH v2] cpufreq: intel_pstate: Change the calculation of next pstate

2014-05-01 Thread Dirk Brandewie
On 05/01/2014 02:00 PM, Stratos Karafotis wrote: Currently the driver calculates the next pstate proportional to core_busy factor, scaled by the ratio max_pstate / current_pstate. Using the scaled load (core_busy) to calculate the next pstate is not always correct, because there are cases that t

[PATCH v2] cpufreq: intel_pstate: Change the calculation of next pstate

2014-05-01 Thread Stratos Karafotis
Currently the driver calculates the next pstate proportional to core_busy factor, scaled by the ratio max_pstate / current_pstate. Using the scaled load (core_busy) to calculate the next pstate is not always correct, because there are cases that the load is independent from current pstate. For ex