[v2,4/4] perf: Tweak perf_event_attr::exclusive semantics
diff mbox series

Message ID 20201029162902.105962225@infradead.org
State Accepted
Commit 1908dc911792067287458fdb0800f036f4f4e0f6
Headers show
Series
  • perf: Fix perf_event_attr::exclusive rotation
Related show

Commit Message

Peter Zijlstra Oct. 29, 2020, 4:27 p.m. UTC
Currently perf_event_attr::exclusive can be used to ensure an
event(group) is the sole group scheduled on the PMU. One consequence
is that when you have a pinned event (say the watchdog) you can no
longer have regular exclusive event(group)s.

Inspired by the fact that !pinned events are considered less strict,
allow !pinned,exclusive events to share the PMU with pinned,!exclusive
events.

Pinned,exclusive is still fully exclusive.

Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
---
 kernel/events/core.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Patch
diff mbox series

--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -2637,7 +2637,7 @@  static int group_can_go_on(struct perf_e
 	 * If this group is exclusive and there are already
 	 * events on the CPU, it can't go on.
 	 */
-	if (event->attr.exclusive && cpuctx->active_oncpu)
+	if (event->attr.exclusive && !list_empty(get_event_list(event)))
 		return 0;
 	/*
 	 * Otherwise, try to add it if all previous groups were able