On 6/24/2021 6:48 AM, Ionela Voinescu wrote: > Not if the counters are implemented properly. The kernel considers that > both reference and delivered performance counters should stop or reset > during idle. The kernel would not account for idle itself. > > If the reference performance counter does not stop during idle, while > the core performance counter (delivered) does stop, the behavior above > should be seen very often. > > Qian, do you see these small delivered performance values often or > seldom? FYI, the latest data point it that on the new kernel, the delivered performance does match the cpufreq_cur_freq. IOW, feedback_ctrs works fine. Also, "powersave" governor could bring down the scaling_cur_freq to scaling_min_freq. Right now, looks like the puzzles on this particular system as mentioned in the previous post are, 1) lowest_freq/lowest_perf != scaling_min_freq 2) CPPC + schedutil is not able to scale down CPUs.
