On Tuesday 17 Sep 2019 at 16:27:46 (+0200), Giovanni Gherdovich wrote: > I'll check what's the cost of static_cpu_has() and if it's non-negligible I'll > do what you suggest (x86-specific version of arch_scale_freq_invariant().
In case this is indeed expensive to check, you could always add a static key, set at boot time, to optimize things a bit ... That might be worth it since this is called in latency-sensitive paths of the scheduler. Thanks, Quentin
