On Wed, Oct 04, 2023 at 05:42:49PM +0200, Ingo Molnar wrote: > > * Namhyung Kim <namhyung@xxxxxxxxxx> wrote: > > > Maybe I can add an option to perf bench sched to place > > senders and receivers in different cgroups. > > That would certainly be useful to measure cgroups overhead. Sent out the change: https://lore.kernel.org/all/20231004204741.985422-1-namhyung@xxxxxxxxxx/ With that, the numbers became stable. :) Before) $ taskset -c 0 ./perf bench sched pipe -l 10000 -G AAA,BBB # Running 'sched/pipe' benchmark: # Executed 10000 pipe operations between two processes Total time: 0.901 [sec] 90.128700 usecs/op 11095 ops/sec After) $ taskset -c 0 ./perf bench sched pipe -l 10000 -G AAA,BBB # Running 'sched/pipe' benchmark: # Executed 10000 pipe operations between two processes Total time: 0.065 [sec] 6.560100 usecs/op 152436 ops/sec Thanks, Namhyung
