Commit 26657848 authored Mar 22, 2016 by Peter Zijlstra Committed by Ingo Molnar Mar 31, 2016

perf/core: Verify we have a single perf_hw_context PMU



There should (and can) only be a single PMU for perf_hw_context
events.

This is because of how we schedule events: once a hardware event fails to
schedule (the PMU is 'full') we stop trying to add more. The trivial
'fix' would break the Round-Robin scheduling we do.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Alexander Shishkin <alexander.shishkin@linux.intel.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Signed-off-by: Ingo Molnar <mingo@kernel.org>

parent 07dc900e

kernel/events/core.c

+9 −0

Original line number	Diff line number	Diff line
		@@ -7693,6 +7693,15 @@ int perf_pmu_register(struct pmu pmu, const char name, int type)
		}

		skip_type:
		if (pmu->task_ctx_nr == perf_hw_context) {
		static int hw_context_taken = 0;

		if (WARN_ON_ONCE(hw_context_taken))
		pmu->task_ctx_nr = perf_invalid_context;

		hw_context_taken = 1;
		}

		pmu->pmu_cpu_context = find_pmu_context(pmu->task_ctx_nr);
		if (pmu->pmu_cpu_context)
		goto got_cpu_context;