Re: [PATCH 1/4] PM/Domains: Add support for retrieving genpd performance states information

[Thara Gopinath <thara.gopinath@linaro.org>]

On 08/22/2019 11:03 AM, Ulf Hansson wrote:
> On Sat, 10 Aug 2019 at 02:58, Thara Gopinath <thara.gopinath@linaro.org> wrote:
>>
>> Add two new APIs in the genpd framework,
>> dev_pm_genpd_get_performance_state to return the current performance
>> state of a power domain and dev_pm_genpd_performance_state_count to
>> return the total number of performance states supported by a
>> power domain. Since the genpd framework does not maintain
>> a count of number of performance states supported by a power domain,
>> introduce a new callback(.get_performance_state_count) that can be used
>> to retrieve this information from power domain drivers.
> 
> I think some brief background to *why* this is useful needs to be
> squeezed into the changelog. Or at least state that following changes
> makes use of it, somehow.
> 
>>
>> Signed-off-by: Thara Gopinath <thara.gopinath@linaro.org>
>> ---
>>  drivers/base/power/domain.c | 38 ++++++++++++++++++++++++++++++++++++++
>>  include/linux/pm_domain.h   | 18 ++++++++++++++++++
>>  2 files changed, 56 insertions(+)
>>
>> diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c
>> index b063bc4..17e0375 100644
>> --- a/drivers/base/power/domain.c
>> +++ b/drivers/base/power/domain.c
>> @@ -413,6 +413,44 @@ int dev_pm_genpd_set_performance_state(struct device *dev, unsigned int state)
>>  }
>>  EXPORT_SYMBOL_GPL(dev_pm_genpd_set_performance_state);
>>
>> +int dev_pm_genpd_get_performance_state(struct device *dev,
>> +                                      unsigned int *state)
>> +{
>> +       struct generic_pm_domain *genpd;
>> +
>> +       genpd = dev_to_genpd(dev);
> 
> We need to verify that the there is a genpd attached before doing this
> cast. Let me post a patch in a day or so, it will give you a helper
> function that covers this.
> 
>> +       if (IS_ERR(genpd))
>> +               return -ENODEV;
>> +
>> +       genpd_lock(genpd);
>> +       *state = genpd->performance_state;
> 
> Why not return the state, rather than assigning an out-parameter?
> 
>> +       genpd_unlock(genpd);
>> +
>> +       return 0;
>> +}
>> +EXPORT_SYMBOL_GPL(dev_pm_genpd_get_performance_state);
>> +
>> +int dev_pm_genpd_performance_state_count(struct device *dev,
>> +                                        unsigned int *count)
>> +{
>> +       struct generic_pm_domain *genpd;
>> +       int ret;
>> +
>> +       genpd = dev_to_genpd(dev);
>> +       if (IS_ERR(genpd))
>> +               return -ENODEV;
>> +
>> +       if (unlikely(!genpd->get_performance_state_count))
>> +               return -EINVAL;
>> +
>> +       genpd_lock(genpd);
>> +       ret = genpd->get_performance_state_count(genpd, count);
> 
> Why not having the callback to return the state, rather than using an
> out-parameter?
Hi Ulf,
I just realized that returning the state instead of using a parameter
will prevent me from access under lock. Is that okay ?

Regards
Thara
> 
>> +       genpd_unlock(genpd);
>> +
>> +       return ret;
>> +}
>> +EXPORT_SYMBOL_GPL(dev_pm_genpd_performance_state_count);
>> +
>>  static int _genpd_power_on(struct generic_pm_domain *genpd, bool timed)
>>  {
>>         unsigned int state_idx = genpd->state_idx;
>> diff --git a/include/linux/pm_domain.h b/include/linux/pm_domain.h
>> index 91d9bf4..0e5f502 100644
>> --- a/include/linux/pm_domain.h
>> +++ b/include/linux/pm_domain.h
>> @@ -117,6 +117,8 @@ struct generic_pm_domain {
>>                                                  struct dev_pm_opp *opp);
>>         int (*set_performance_state)(struct generic_pm_domain *genpd,
>>                                      unsigned int state);
>> +       int (*get_performance_state_count)(struct generic_pm_domain *genpd,
>> +                                          unsigned int *count);
>>         struct gpd_dev_ops dev_ops;
>>         s64 max_off_time_ns;    /* Maximum allowed "suspended" time. */
>>         bool max_off_time_changed;
>> @@ -204,6 +206,10 @@ int pm_genpd_init(struct generic_pm_domain *genpd,
>>                   struct dev_power_governor *gov, bool is_off);
>>  int pm_genpd_remove(struct generic_pm_domain *genpd);
>>  int dev_pm_genpd_set_performance_state(struct device *dev, unsigned int state);
>> +int dev_pm_genpd_get_performance_state(struct device *dev,
>> +                                      unsigned int *state);
>> +int dev_pm_genpd_performance_state_count(struct device *dev,
>> +                                        unsigned int *count);
>>
>>  extern struct dev_power_governor simple_qos_governor;
>>  extern struct dev_power_governor pm_domain_always_on_gov;
>> @@ -251,6 +257,18 @@ static inline int dev_pm_genpd_set_performance_state(struct device *dev,
>>         return -ENOTSUPP;
>>  }
>>
>> +static inline int dev_pm_genpd_get_performance_state(struct device *dev,
>> +                                                    unsigned int *state);
>> +{
>> +       return -ENOTSUPP;
>> +}
>> +
>> +static inline int dev_pm_genpd_performance_state_count(struct device *dev,
>> +                                                      unsigned int *count);
>> +{
>> +       return -ENOTSUPP;
>> +}
>> +
>>  #define simple_qos_governor            (*(struct dev_power_governor *)(NULL))
>>  #define pm_domain_always_on_gov                (*(struct dev_power_governor *)(NULL))
>>  #endif
>> --
>> 2.1.4
>>
> 
> Kind regards
> Uffe
> 


-- 
Regards
Thara