Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit a37e5be2 authored by Raghavendra Rao Ananta's avatar Raghavendra Rao Ananta
Browse files

perf: Cancel the mux hrtimer during CPU hotplug to avoid migration



The current design of hrtimers migrates the pinned timers to a
different CPU upon its hotplug. However, perf-core needs to
maintain the mux-hrtimers on a per CPU basis. That is, each
hrtimer carries the context for that particular CPU and would
lose this context if it gets migrated to a different CPU. As a
result, cancel the hrtimer for the CPU that's about to go down
and restart it (if required) when the perf-events are being created.

Change-Id: I7a1d0456208855e3a99a7d49e59c6dae811d146e
Signed-off-by: default avatarRaghavendra Rao Ananta <rananta@codeaurora.org>
parent 77257e46
Loading
Loading
Loading
Loading
+11 −0
Original line number Diff line number Diff line
@@ -11532,6 +11532,7 @@ static void perf_event_exit_cpu_context(int cpu)
	struct perf_cpu_context *cpuctx;
	struct perf_event_context *ctx;
	struct perf_event *event, *event_tmp;
	unsigned long flags;
	struct pmu *pmu;

	mutex_lock(&pmus_lock);
@@ -11539,6 +11540,16 @@ static void perf_event_exit_cpu_context(int cpu)
	list_for_each_entry(pmu, &pmus, entry) {
		cpuctx = per_cpu_ptr(pmu->pmu_cpu_context, cpu);
		ctx = &cpuctx->ctx;

		/* Cancel the mux hrtimer to avoid CPU migration */
		if (pmu->task_ctx_nr != perf_sw_context) {
			raw_spin_lock_irqsave(&cpuctx->hrtimer_lock, flags);
			hrtimer_cancel(&cpuctx->hrtimer);
			cpuctx->hrtimer_active = 0;
			raw_spin_unlock_irqrestore(&cpuctx->hrtimer_lock,
							flags);
		}

		mutex_lock(&ctx->mutex);
		list_for_each_entry_safe(event, event_tmp, &ctx->event_list,
								event_entry) {