On Monday 20 Jan 2020 at 16:20:49 (+0000), Lukasz Luba wrote: > On 1/20/20 3:28 PM, Quentin Perret wrote: > > Agreed, this looks a bit confusing. It should be trivial to make > > em_dev_get() (or whatever we end up calling it) work for CPUs too, > > though. And we could always have a em_cpu_get(int cpu) API that is a > > basically a wrapper around em_dev_get() for convenience. > > The problem not only here is that we have a CPU index 'int cpu' > and if we ask for device like: > > struct device *dev = get_cpu_device(cpu); > > It might be not the same device that was used during the > registration, when we had i.e. 4 CPUs for the same policy: > > int cpu_id = cpumask_first(policy->cpus); > struct device *cpu_dev = get_cpu_device(cpu_id); > em_register_perf_domain(cpu_dev, nr_opp, &em_cb); > > That's why the em_cpu_get() is different than em_get_pd(), mainly by: > if (cpumask_test_cpu(cpu, em_span_cpus(em_pd))) > > It won't be simple wrapper, let me think how it could be handled > differently than it is now. Right so I suppose the easiest solution would be to do the opposite of my first suggestion. That is, make em_get_pd() call em_cpu_get() if the device is a CPU device, or proceed to the PD list iteration for other devices. And em_cpu_get() can remain as you originally suggested (that is, iterate over the PDs and test the mask). That should ensure em_get_pd() always works, em_cpu_get() is still there handy for the scheduler and such, and the two EM lookup functions (for CPUs or for devices) are kept cleanly separated. Thoughts ? Thanks, Quentin
