On Wed, Sep 23, 2020 at 5:14 PM Rafael J. Wysocki <[email protected]> wrote: > > On Wed, Sep 16, 2020 at 8:46 AM Viresh Kumar <[email protected]> wrote: > > > > Now that all the blockers are gone for enabling stats in fast-switching > > case, enable it. > > > > Signed-off-by: Viresh Kumar <[email protected]> > > --- > > drivers/cpufreq/cpufreq.c | 6 +++++- > > drivers/cpufreq/cpufreq_stats.c | 6 ------ > > 2 files changed, 5 insertions(+), 7 deletions(-) > > > > diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c > > index 47aa90f9a7c2..d5fe64e96be9 100644 > > --- a/drivers/cpufreq/cpufreq.c > > +++ b/drivers/cpufreq/cpufreq.c > > @@ -2057,8 +2057,12 @@ unsigned int cpufreq_driver_fast_switch(struct > > cpufreq_policy *policy, > > unsigned int target_freq) > > { > > target_freq = clamp_val(target_freq, policy->min, policy->max); > > + target_freq = cpufreq_driver->fast_switch(policy, target_freq); > > > > - return cpufreq_driver->fast_switch(policy, target_freq); > > + if (target_freq) > > + cpufreq_stats_record_transition(policy, target_freq); > > So this adds two extra branches in the scheduler path for the cases > when the stats are not used at all which seems avoidable to some > extent. > > Can we check policy->stats upfront here and bail out right away if it > is not set, for example?
Well, scratch this, the next patch fixes it up. Cheers!
