| Message ID | 20210602101215.78094-3-ulf.hansson@linaro.org |
|---|---|
| State | New |
| Headers | show |
| Series | PM: domains: Avoid boilerplate code for DVFS in subsystem/drivers | expand |
On 02-06-21, 12:12, Ulf Hansson wrote: > A subsystem/driver that need to manage OPPs for its device, should > typically drop its vote for the OPP when the device becomes runtime > suspended. In this way, the corresponding aggregation of the performance > state votes that is managed in genpd for the attached PM domain, may find > that the aggregated vote can be decreased. Hence, it may allow genpd to set > the lower performance state for the PM domain, thus avoiding to waste > energy. > > To accomplish this, typically a subsystem/driver would need to call > dev_pm_opp_set_rate|opp() for its device from its ->runtime_suspend() > callback, to drop the vote for the OPP. Accordingly, it needs another call > to dev_pm_opp_set_rate|opp() to restore the vote for the OPP from its > ->runtime_resume() callback. > > To avoid boilerplate code in subsystems/driver to deal with these things, > let's instead manage this internally in genpd. > > Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org> > --- > drivers/base/power/domain.c | 21 +++++++++++++++++++-- > include/linux/pm_domain.h | 1 + > 2 files changed, 20 insertions(+), 2 deletions(-) > > diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c > index a3b6e751f366..81b9d4652b90 100644 > --- a/drivers/base/power/domain.c > +++ b/drivers/base/power/domain.c > @@ -397,6 +397,18 @@ static int genpd_set_performance_state(struct device *dev, unsigned int state) > return ret; > } > > +static int genpd_drop_performance_state(struct device *dev) > +{ > + struct generic_pm_domain_data *gpd_data = dev_gpd_data(dev); > + unsigned int prev_state; > + > + prev_state = gpd_data->performance_state; > + if (prev_state && !genpd_set_performance_state(dev, 0)) What about adding this prev_state check in genpd_set_performance_state() itself ? We already have one for the genpd in _genpd_set_performance_state(), why not one for the device ? > + return prev_state; > + > + return 0; Hmm, we will return 0 in case genpd_set_performance_state() fails, which will make us set the state to 0 again on resume. Maybe add a comment for this somewhere ? > +} > + > /** > * dev_pm_genpd_set_performance_state- Set performance state of device's power > * domain. > @@ -839,7 +851,8 @@ static int genpd_runtime_suspend(struct device *dev) > { > struct generic_pm_domain *genpd; > bool (*suspend_ok)(struct device *__dev); > - struct gpd_timing_data *td = &dev_gpd_data(dev)->td; > + struct generic_pm_domain_data *gpd_data = dev_gpd_data(dev); > + struct gpd_timing_data *td = &gpd_data->td; > bool runtime_pm = pm_runtime_enabled(dev); > ktime_t time_start; > s64 elapsed_ns; > @@ -896,6 +909,7 @@ static int genpd_runtime_suspend(struct device *dev) > return 0; > > genpd_lock(genpd); > + gpd_data->rpm_saved_pstate = genpd_drop_performance_state(dev); > genpd_power_off(genpd, true, 0); > genpd_unlock(genpd); > > @@ -913,7 +927,8 @@ static int genpd_runtime_suspend(struct device *dev) > static int genpd_runtime_resume(struct device *dev) > { > struct generic_pm_domain *genpd; > - struct gpd_timing_data *td = &dev_gpd_data(dev)->td; > + struct generic_pm_domain_data *gpd_data = dev_gpd_data(dev); > + struct gpd_timing_data *td = &gpd_data->td; > bool runtime_pm = pm_runtime_enabled(dev); > ktime_t time_start; > s64 elapsed_ns; > @@ -937,6 +952,8 @@ static int genpd_runtime_resume(struct device *dev) > > genpd_lock(genpd); > ret = genpd_power_on(genpd, 0); > + if (!ret && gpd_data->rpm_saved_pstate) > + genpd_set_performance_state(dev, gpd_data->rpm_saved_pstate); > genpd_unlock(genpd); > > if (ret) > diff --git a/include/linux/pm_domain.h b/include/linux/pm_domain.h > index dfcfbcecc34b..c3d6c15788a3 100644 > --- a/include/linux/pm_domain.h > +++ b/include/linux/pm_domain.h > @@ -198,6 +198,7 @@ struct generic_pm_domain_data { > struct notifier_block *power_nb; > int cpu; > unsigned int performance_state; > + unsigned int rpm_saved_pstate; > ktime_t next_wakeup; > void *data; > }; > -- > 2.25.1 -- viresh
On 03-06-21, 08:04, Viresh Kumar wrote: > On 02-06-21, 12:12, Ulf Hansson wrote: > > A subsystem/driver that need to manage OPPs for its device, should > > typically drop its vote for the OPP when the device becomes runtime > > suspended. In this way, the corresponding aggregation of the performance > > state votes that is managed in genpd for the attached PM domain, may find > > that the aggregated vote can be decreased. Hence, it may allow genpd to set > > the lower performance state for the PM domain, thus avoiding to waste > > energy. > > > > To accomplish this, typically a subsystem/driver would need to call > > dev_pm_opp_set_rate|opp() for its device from its ->runtime_suspend() > > callback, to drop the vote for the OPP. Accordingly, it needs another call > > to dev_pm_opp_set_rate|opp() to restore the vote for the OPP from its > > ->runtime_resume() callback. > > > > To avoid boilerplate code in subsystems/driver to deal with these things, > > let's instead manage this internally in genpd. > > > > Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org> > > --- > > drivers/base/power/domain.c | 21 +++++++++++++++++++-- > > include/linux/pm_domain.h | 1 + > > 2 files changed, 20 insertions(+), 2 deletions(-) > > > > diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c > > index a3b6e751f366..81b9d4652b90 100644 > > --- a/drivers/base/power/domain.c > > +++ b/drivers/base/power/domain.c > > @@ -397,6 +397,18 @@ static int genpd_set_performance_state(struct device *dev, unsigned int state) > > return ret; > > } > > > > +static int genpd_drop_performance_state(struct device *dev) > > +{ > > + struct generic_pm_domain_data *gpd_data = dev_gpd_data(dev); > > + unsigned int prev_state; > > + > > + prev_state = gpd_data->performance_state; > > + if (prev_state && !genpd_set_performance_state(dev, 0)) > > What about adding this prev_state check in > genpd_set_performance_state() itself ? We already have one for the > genpd in _genpd_set_performance_state(), why not one for the device ? > > > + return prev_state; > > + > > + return 0; > > Hmm, we will return 0 in case genpd_set_performance_state() fails, > which will make us set the state to 0 again on resume. Maybe add a > comment for this somewhere ? No, we won't as you check for rpm_saved_pstate there, so the device will stay disabled. Again adding the check into genpd_set_performance_state() may help reducing similar checks elsewhere. > > +} > > + > > /** > > * dev_pm_genpd_set_performance_state- Set performance state of device's power > > * domain. > > @@ -839,7 +851,8 @@ static int genpd_runtime_suspend(struct device *dev) > > { > > struct generic_pm_domain *genpd; > > bool (*suspend_ok)(struct device *__dev); > > - struct gpd_timing_data *td = &dev_gpd_data(dev)->td; > > + struct generic_pm_domain_data *gpd_data = dev_gpd_data(dev); > > + struct gpd_timing_data *td = &gpd_data->td; > > bool runtime_pm = pm_runtime_enabled(dev); > > ktime_t time_start; > > s64 elapsed_ns; > > @@ -896,6 +909,7 @@ static int genpd_runtime_suspend(struct device *dev) > > return 0; > > > > genpd_lock(genpd); > > + gpd_data->rpm_saved_pstate = genpd_drop_performance_state(dev); > > genpd_power_off(genpd, true, 0); > > genpd_unlock(genpd); > > > > @@ -913,7 +927,8 @@ static int genpd_runtime_suspend(struct device *dev) > > static int genpd_runtime_resume(struct device *dev) > > { > > struct generic_pm_domain *genpd; > > - struct gpd_timing_data *td = &dev_gpd_data(dev)->td; > > + struct generic_pm_domain_data *gpd_data = dev_gpd_data(dev); > > + struct gpd_timing_data *td = &gpd_data->td; > > bool runtime_pm = pm_runtime_enabled(dev); > > ktime_t time_start; > > s64 elapsed_ns; > > @@ -937,6 +952,8 @@ static int genpd_runtime_resume(struct device *dev) > > > > genpd_lock(genpd); > > ret = genpd_power_on(genpd, 0); > > + if (!ret && gpd_data->rpm_saved_pstate) > > + genpd_set_performance_state(dev, gpd_data->rpm_saved_pstate); > > genpd_unlock(genpd); -- viresh
On Thu, 3 Jun 2021 at 04:37, Viresh Kumar <viresh.kumar@linaro.org> wrote: > > On 03-06-21, 08:04, Viresh Kumar wrote: > > On 02-06-21, 12:12, Ulf Hansson wrote: > > > A subsystem/driver that need to manage OPPs for its device, should > > > typically drop its vote for the OPP when the device becomes runtime > > > suspended. In this way, the corresponding aggregation of the performance > > > state votes that is managed in genpd for the attached PM domain, may find > > > that the aggregated vote can be decreased. Hence, it may allow genpd to set > > > the lower performance state for the PM domain, thus avoiding to waste > > > energy. > > > > > > To accomplish this, typically a subsystem/driver would need to call > > > dev_pm_opp_set_rate|opp() for its device from its ->runtime_suspend() > > > callback, to drop the vote for the OPP. Accordingly, it needs another call > > > to dev_pm_opp_set_rate|opp() to restore the vote for the OPP from its > > > ->runtime_resume() callback. > > > > > > To avoid boilerplate code in subsystems/driver to deal with these things, > > > let's instead manage this internally in genpd. > > > > > > Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org> > > > --- > > > drivers/base/power/domain.c | 21 +++++++++++++++++++-- > > > include/linux/pm_domain.h | 1 + > > > 2 files changed, 20 insertions(+), 2 deletions(-) > > > > > > diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c > > > index a3b6e751f366..81b9d4652b90 100644 > > > --- a/drivers/base/power/domain.c > > > +++ b/drivers/base/power/domain.c > > > @@ -397,6 +397,18 @@ static int genpd_set_performance_state(struct device *dev, unsigned int state) > > > return ret; > > > } > > > > > > +static int genpd_drop_performance_state(struct device *dev) > > > +{ > > > + struct generic_pm_domain_data *gpd_data = dev_gpd_data(dev); > > > + unsigned int prev_state; > > > + > > > + prev_state = gpd_data->performance_state; > > > + if (prev_state && !genpd_set_performance_state(dev, 0)) > > > > What about adding this prev_state check in > > genpd_set_performance_state() itself ? We already have one for the > > genpd in _genpd_set_performance_state(), why not one for the device ? > > > > > + return prev_state; > > > + > > > + return 0; > > > > Hmm, we will return 0 in case genpd_set_performance_state() fails, > > which will make us set the state to 0 again on resume. Maybe add a > > comment for this somewhere ? > > No, we won't as you check for rpm_saved_pstate there, so the device > will stay disabled. > > Again adding the check into genpd_set_performance_state() may help > reducing similar checks elsewhere. Yes, at closer look, I think it makes sense to me as well. Although, as it means a change in behaviour, I decided to make it a separate patch. Let me respin the series to fold it in. [...] Thanks for reviewing! Kind regards Uffe
diff --git a/drivers/base/power/domain.c b/drivers/base/power/domain.c index a3b6e751f366..81b9d4652b90 100644 --- a/drivers/base/power/domain.c +++ b/drivers/base/power/domain.c @@ -397,6 +397,18 @@ static int genpd_set_performance_state(struct device *dev, unsigned int state) return ret; } +static int genpd_drop_performance_state(struct device *dev) +{ + struct generic_pm_domain_data *gpd_data = dev_gpd_data(dev); + unsigned int prev_state; + + prev_state = gpd_data->performance_state; + if (prev_state && !genpd_set_performance_state(dev, 0)) + return prev_state; + + return 0; +} + /** * dev_pm_genpd_set_performance_state- Set performance state of device's power * domain. @@ -839,7 +851,8 @@ static int genpd_runtime_suspend(struct device *dev) { struct generic_pm_domain *genpd; bool (*suspend_ok)(struct device *__dev); - struct gpd_timing_data *td = &dev_gpd_data(dev)->td; + struct generic_pm_domain_data *gpd_data = dev_gpd_data(dev); + struct gpd_timing_data *td = &gpd_data->td; bool runtime_pm = pm_runtime_enabled(dev); ktime_t time_start; s64 elapsed_ns; @@ -896,6 +909,7 @@ static int genpd_runtime_suspend(struct device *dev) return 0; genpd_lock(genpd); + gpd_data->rpm_saved_pstate = genpd_drop_performance_state(dev); genpd_power_off(genpd, true, 0); genpd_unlock(genpd); @@ -913,7 +927,8 @@ static int genpd_runtime_suspend(struct device *dev) static int genpd_runtime_resume(struct device *dev) { struct generic_pm_domain *genpd; - struct gpd_timing_data *td = &dev_gpd_data(dev)->td; + struct generic_pm_domain_data *gpd_data = dev_gpd_data(dev); + struct gpd_timing_data *td = &gpd_data->td; bool runtime_pm = pm_runtime_enabled(dev); ktime_t time_start; s64 elapsed_ns; @@ -937,6 +952,8 @@ static int genpd_runtime_resume(struct device *dev) genpd_lock(genpd); ret = genpd_power_on(genpd, 0); + if (!ret && gpd_data->rpm_saved_pstate) + genpd_set_performance_state(dev, gpd_data->rpm_saved_pstate); genpd_unlock(genpd); if (ret) diff --git a/include/linux/pm_domain.h b/include/linux/pm_domain.h index dfcfbcecc34b..c3d6c15788a3 100644 --- a/include/linux/pm_domain.h +++ b/include/linux/pm_domain.h @@ -198,6 +198,7 @@ struct generic_pm_domain_data { struct notifier_block *power_nb; int cpu; unsigned int performance_state; + unsigned int rpm_saved_pstate; ktime_t next_wakeup; void *data; };
A subsystem/driver that need to manage OPPs for its device, should typically drop its vote for the OPP when the device becomes runtime suspended. In this way, the corresponding aggregation of the performance state votes that is managed in genpd for the attached PM domain, may find that the aggregated vote can be decreased. Hence, it may allow genpd to set the lower performance state for the PM domain, thus avoiding to waste energy. To accomplish this, typically a subsystem/driver would need to call dev_pm_opp_set_rate|opp() for its device from its ->runtime_suspend() callback, to drop the vote for the OPP. Accordingly, it needs another call to dev_pm_opp_set_rate|opp() to restore the vote for the OPP from its ->runtime_resume() callback. To avoid boilerplate code in subsystems/driver to deal with these things, let's instead manage this internally in genpd. Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org> --- drivers/base/power/domain.c | 21 +++++++++++++++++++-- include/linux/pm_domain.h | 1 + 2 files changed, 20 insertions(+), 2 deletions(-) -- 2.25.1