perf sched: Sanity check context switch events
authorIngo Molnar <mingo@elte.hu>
Wed, 16 Sep 2009 12:07:00 +0000 (14:07 +0200)
committerIngo Molnar <mingo@elte.hu>
Wed, 16 Sep 2009 10:08:28 +0000 (12:08 +0200)
Use 'perf sched latency' to track the current task based on
context-switch events, and flag the cases where there's some
impossible transition: such as a PID being switched out that
was not switched in.

Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
tools/perf/builtin-sched.c

index 1f0f9be..2d54236 100644 (file)
@@ -119,6 +119,7 @@ static unsigned long                replay_repeat = 10;
 static unsigned long           nr_timestamps;
 static unsigned long           nr_unordered_timestamps;
 static unsigned long           nr_state_machine_bugs;
+static unsigned long           nr_context_switch_bugs;
 static unsigned long           nr_events;
 static unsigned long           nr_lost_chunks;
 static unsigned long           nr_lost_events;
@@ -1399,6 +1400,14 @@ static void __cmd_lat(void)
                        printf(" (due to lost events?)");
                printf("\n");
        }
+       if (nr_context_switch_bugs && nr_timestamps) {
+               printf("  INFO: %.3f%% context switch bugs (%ld out of %ld)",
+                       (double)nr_context_switch_bugs/(double)nr_timestamps*100.0,
+                       nr_context_switch_bugs, nr_timestamps);
+               if (nr_lost_events)
+                       printf(" (due to lost events?)");
+               printf("\n");
+       }
        printf("\n");
 
 }
@@ -1425,10 +1434,16 @@ process_sched_wakeup_event(struct raw_event_sample *raw,
        trace_handler->wakeup_event(&wakeup_event, event, cpu, timestamp, thread);
 }
 
+/*
+ * Track the current task - that way we can know whether there's any
+ * weird events, such as a task being switched away that is not current.
+ */
+static u32 curr_pid[MAX_CPUS] = { [0 ... MAX_CPUS-1] = -1 };
+
 static void
 process_sched_switch_event(struct raw_event_sample *raw,
                           struct event *event,
-                          int cpu __used,
+                          int cpu,
                           u64 timestamp __used,
                           struct thread *thread __used)
 {
@@ -1444,6 +1459,16 @@ process_sched_switch_event(struct raw_event_sample *raw,
        FILL_FIELD(switch_event, next_pid, event, raw->data);
        FILL_FIELD(switch_event, next_prio, event, raw->data);
 
+       if (curr_pid[cpu] != (u32)-1) {
+               /*
+                * Are we trying to switch away a PID that is
+                * not current?
+                */
+               if (curr_pid[cpu] != switch_event.prev_pid)
+                       nr_context_switch_bugs++;
+       }
+       curr_pid[cpu] = switch_event.next_pid;
+
        trace_handler->switch_event(&switch_event, event, cpu, timestamp, thread);
 }