On Wednesday 15 May 2019 at 12:54:10 (+0200), Daniel Lezcano wrote: > On 15/05/2019 12:40, Quentin Perret wrote: > > On Wednesday 15 May 2019 at 15:52:00 (+0530), Viresh Kumar wrote: > >> On 15-05-19, 12:16, Daniel Lezcano wrote: > >>> Viresh what do you think ? > >> > >> I agree with your last suggestions. They do make sense. > > > > Good :-) > > > > So, FWIW, the below compiles w/ or w/o THERMAL_GOV_POWER_ALLOCATOR. I'll > > test it and clean it up some more and put it as patch 1 in the series if > > that's OK. > > > > Thanks, > > Quentin > > > > > > diff --git a/drivers/thermal/cpu_cooling.c b/drivers/thermal/cpu_cooling.c > > index f7c1f49ec87f..ee431848ef71 100644 > > --- a/drivers/thermal/cpu_cooling.c > > +++ b/drivers/thermal/cpu_cooling.c > > @@ -58,7 +58,9 @@ > > */ > > struct freq_table { > > u32 frequency; > > I suspect we will have a problem here as cpufreq_set_cur_state uses the > frequency and when switching to EM, we will still need a freq table.
Indeed, I'll need a bit of ifdefery in the get_state_freq() function introduced in patch 3, but nothing too horrible I hope. Thanks, Quentin