perf/x86/intel/uncore: Allow only a single PMU/box within an events group
authorPeter Zijlstra <peterz@infradead.org>
Fri, 18 Nov 2016 12:53:54 +0000 (13:53 +0100)
committerIngo Molnar <mingo@kernel.org>
Tue, 22 Nov 2016 11:36:59 +0000 (12:36 +0100)
Group validation expects all events to be of the same PMU; however
is_uncore_pmu() is too wide, it matches _all_ uncore events, even
across PMUs.

This triggers failure when we group different events from different
uncore PMUs, like:

  perf stat -vv -e '{uncore_cbox_0/config=0x0334/,uncore_qpi_0/event=1/}' -a sleep 1

Fix is_uncore_pmu() by only matching events to the box at hand.

Note that generic code; ran after this step; will disallow this
mixture of PMU events.

Reported-by: Jiri Olsa <jolsa@redhat.com>
Tested-by: Jiri Olsa <jolsa@redhat.com>
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: Kan Liang <kan.liang@intel.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 <vince@deater.net>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Link: http://lkml.kernel.org/r/20161118125354.GQ3117@twins.programming.kicks-ass.net
Signed-off-by: Ingo Molnar <mingo@kernel.org>
arch/x86/events/intel/uncore.c

index efca2685d8767582e624d1c38779e5c96eef33b9..dbaaf7dc8373cb0248a4637abe4a211f266fb2db 100644 (file)
@@ -319,9 +319,9 @@ static struct intel_uncore_box *uncore_alloc_box(struct intel_uncore_type *type,
  */
 static int uncore_pmu_event_init(struct perf_event *event);
 
-static bool is_uncore_event(struct perf_event *event)
+static bool is_box_event(struct intel_uncore_box *box, struct perf_event *event)
 {
-       return event->pmu->event_init == uncore_pmu_event_init;
+       return &box->pmu->pmu == event->pmu;
 }
 
 static int
@@ -340,7 +340,7 @@ uncore_collect_events(struct intel_uncore_box *box, struct perf_event *leader,
 
        n = box->n_events;
 
-       if (is_uncore_event(leader)) {
+       if (is_box_event(box, leader)) {
                box->event_list[n] = leader;
                n++;
        }
@@ -349,7 +349,7 @@ uncore_collect_events(struct intel_uncore_box *box, struct perf_event *leader,
                return n;
 
        list_for_each_entry(event, &leader->sibling_list, group_entry) {
-               if (!is_uncore_event(event) ||
+               if (!is_box_event(box, event) ||
                    event->state <= PERF_EVENT_STATE_OFF)
                        continue;