perf_events: Revert: Fix transaction recovery in group_sched_in()
authorStephane Eranian <eranian@google.com>
Wed, 20 Oct 2010 13:25:01 +0000 (15:25 +0200)
committerIngo Molnar <mingo@elte.hu>
Fri, 22 Oct 2010 12:18:27 +0000 (14:18 +0200)
This patch reverts commit 8e5fc1a (perf_events: Fix transaction
recovery in group_sched_in()) because it had one flaw in case the
group could never be scheduled. It would cause time_enabled to get
negative.

Signed-off-by: Stephane Eranian <eranian@google.com>
Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
LKML-Reference: <4cbeeeb7.0aefd80a.6e40.0e2f@mx.google.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
kernel/perf_event.c

index f309e8014c7853105d1a38bc662f10164dc4d3d1..39afdb07d75810b9cb26da3c1c5e778adae43674 100644 (file)
@@ -417,8 +417,8 @@ event_filter_match(struct perf_event *event)
        return event->cpu == -1 || event->cpu == smp_processor_id();
 }
 
-static int
-__event_sched_out(struct perf_event *event,
+static void
+event_sched_out(struct perf_event *event,
                  struct perf_cpu_context *cpuctx,
                  struct perf_event_context *ctx)
 {
@@ -437,13 +437,14 @@ __event_sched_out(struct perf_event *event,
        }
 
        if (event->state != PERF_EVENT_STATE_ACTIVE)
-               return 0;
+               return;
 
        event->state = PERF_EVENT_STATE_INACTIVE;
        if (event->pending_disable) {
                event->pending_disable = 0;
                event->state = PERF_EVENT_STATE_OFF;
        }
+       event->tstamp_stopped = ctx->time;
        event->pmu->del(event, 0);
        event->oncpu = -1;
 
@@ -452,19 +453,6 @@ __event_sched_out(struct perf_event *event,
        ctx->nr_active--;
        if (event->attr.exclusive || !cpuctx->active_oncpu)
                cpuctx->exclusive = 0;
-       return 1;
-}
-
-static void
-event_sched_out(struct perf_event *event,
-                 struct perf_cpu_context *cpuctx,
-                 struct perf_event_context *ctx)
-{
-       int ret;
-
-       ret = __event_sched_out(event, cpuctx, ctx);
-       if (ret)
-               event->tstamp_stopped = ctx->time;
 }
 
 static void
@@ -664,7 +652,7 @@ retry:
 }
 
 static int
-__event_sched_in(struct perf_event *event,
+event_sched_in(struct perf_event *event,
                 struct perf_cpu_context *cpuctx,
                 struct perf_event_context *ctx)
 {
@@ -684,6 +672,8 @@ __event_sched_in(struct perf_event *event,
                return -EAGAIN;
        }
 
+       event->tstamp_running += ctx->time - event->tstamp_stopped;
+
        if (!is_software_event(event))
                cpuctx->active_oncpu++;
        ctx->nr_active++;
@@ -694,35 +684,6 @@ __event_sched_in(struct perf_event *event,
        return 0;
 }
 
-static inline int
-event_sched_in(struct perf_event *event,
-                struct perf_cpu_context *cpuctx,
-                struct perf_event_context *ctx)
-{
-       int ret = __event_sched_in(event, cpuctx, ctx);
-       if (ret)
-               return ret;
-       event->tstamp_running += ctx->time - event->tstamp_stopped;
-       return 0;
-}
-
-static void
-group_commit_event_sched_in(struct perf_event *group_event,
-              struct perf_cpu_context *cpuctx,
-              struct perf_event_context *ctx)
-{
-       struct perf_event *event;
-       u64 now = ctx->time;
-
-       group_event->tstamp_running += now - group_event->tstamp_stopped;
-       /*
-        * Schedule in siblings as one group (if any):
-        */
-       list_for_each_entry(event, &group_event->sibling_list, group_entry) {
-               event->tstamp_running += now - event->tstamp_stopped;
-       }
-}
-
 static int
 group_sched_in(struct perf_event *group_event,
               struct perf_cpu_context *cpuctx,
@@ -736,13 +697,7 @@ group_sched_in(struct perf_event *group_event,
 
        pmu->start_txn(pmu);
 
-       /*
-        * use __event_sched_in() to delay updating tstamp_running
-        * until the transaction is committed. In case of failure
-        * we will keep an unmodified tstamp_running which is a
-        * requirement to get correct timing information
-        */
-       if (__event_sched_in(group_event, cpuctx, ctx)) {
+       if (event_sched_in(group_event, cpuctx, ctx)) {
                pmu->cancel_txn(pmu);
                return -EAGAIN;
        }
@@ -751,31 +706,26 @@ group_sched_in(struct perf_event *group_event,
         * Schedule in siblings as one group (if any):
         */
        list_for_each_entry(event, &group_event->sibling_list, group_entry) {
-               if (__event_sched_in(event, cpuctx, ctx)) {
+               if (event_sched_in(event, cpuctx, ctx)) {
                        partial_group = event;
                        goto group_error;
                }
        }
 
-       if (!pmu->commit_txn(pmu)) {
-               /* commit tstamp_running */
-               group_commit_event_sched_in(group_event, cpuctx, ctx);
+       if (!pmu->commit_txn(pmu))
                return 0;
-       }
+
 group_error:
        /*
         * Groups can be scheduled in as one unit only, so undo any
         * partial group before returning:
-        *
-        * use __event_sched_out() to avoid updating tstamp_stopped
-        * because the event never actually ran
         */
        list_for_each_entry(event, &group_event->sibling_list, group_entry) {
                if (event == partial_group)
                        break;
-               __event_sched_out(event, cpuctx, ctx);
+               event_sched_out(event, cpuctx, ctx);
        }
-       __event_sched_out(group_event, cpuctx, ctx);
+       event_sched_out(group_event, cpuctx, ctx);
 
        pmu->cancel_txn(pmu);