perf: Provide a separate task context for swevents

Since software events are always schedulable, mixing them up with
hardware events (who are not) can lead to funny scheduling oddities.

Giving them their own context solves this.

Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: paulus <paulus@samba.org>
Cc: stephane eranian <eranian@googlemail.com>
Cc: Robert Richter <robert.richter@amd.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Lin Ming <ming.m.lin@intel.com>
Cc: Yanmin <yanmin_zhang@linux.intel.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>

2 files changed, 2 insertions, 8 deletions

diff --git a/include/linux/perf_event.h b/include/linux/perf_event.h
index 9ecfd856ce6e..c1173520f14d 100644
--- a/include/linux/perf_event.h
+++ b/include/linux/perf_event.h
@@ -952,14 +952,7 @@ extern int perf_event_overflow(struct perf_event *event, int nmi,
  */
 static inline int is_software_event(struct perf_event *event)
 {
-        switch (event->attr.type) {
-        case PERF_TYPE_SOFTWARE:
-        case PERF_TYPE_TRACEPOINT:
-        /* for now the breakpoint stuff also works as software event */
-        case PERF_TYPE_BREAKPOINT:
-                return 1;
-        }
-        return 0;
+        return event->pmu->task_ctx_nr == perf_sw_context;
 }
 
 extern atomic_t perf_swevent_enabled[PERF_COUNT_SW_MAX];
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 89d6023c6f82..eb3c1ceec06e 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1163,6 +1163,7 @@ struct rcu_node;
 enum perf_event_task_context {
         perf_invalid_context = -1,
         perf_hw_context = 0,
+        perf_sw_context,
         perf_nr_task_contexts,
 };