On Mon, Apr 09, 2018 at 10:25:32AM +0300, Alexey Budankov wrote: > > Store preempting context switch out event into Perf trace as a part of > PERF_RECORD_SWITCH[_CPU_WIDE] record. > > Percentage of preempting and non-preempting context switches help > understanding the nature of workloads (CPU or IO bound) that are running > on a machine; > > The event is treated as preemption one when task->state value of the > thread being switched out is TASK_RUNNING. Event type encoding is > implemented using PERF_RECORD_MISC_SWITCH_OUT_PREEMPT bit; > > Signed-off-by: Alexey Budankov <[email protected]>
Acked-by: Peter Zijlstra (Intel) <[email protected]> Acme, I'm thinking you should route this, since most of the changes are actually to the tool. > diff --git a/kernel/events/core.c b/kernel/events/core.c > index fc1c330c6bd6..872a5aaa77eb 100644 > --- a/kernel/events/core.c > +++ b/kernel/events/core.c > @@ -7584,6 +7584,10 @@ static void perf_event_switch(struct task_struct *task, > }, > }; > > + if (!sched_in && task->state == TASK_RUNNING) > + switch_event.event_id.header.misc |= > + PERF_RECORD_MISC_SWITCH_OUT_PREEMPT; I typically prefer {} over any multi-line expression. > perf_iterate_sb(perf_event_switch_output, > &switch_event, > NULL);
