Revert "perf: Do not allow optimized switch for non-cloned events"
authorJiri Olsa <jolsa@redhat.com>
Fri, 12 Sep 2014 11:18:28 +0000 (13:18 +0200)
committerIngo Molnar <mingo@kernel.org>
Wed, 24 Sep 2014 12:48:13 +0000 (14:48 +0200)
This reverts commit 1f9a7268c67f0290837aada443d28fd953ddca90.

With the fix of the initial state for the cloned event we now correctly
handle the error described in:

  1f9a7268c67f perf: Do not allow optimized switch for non-cloned events

so we can revert it.

I made an automated test for this, but its not suitable for automated
perf tests framework. It needs to be customized for each machine (the
more cpu the higher numbers for GROUPS/WORKERS/BYTES) and it could take
longer time to hit the issue.

Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Stephane Eranian <eranian@google.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/20140910143535.GD2409@krava.brq.redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
kernel/events/core.c

index 132524c8b34037de070fb65829642d7829157544..b164cb07b30df175cb10398794a2cea0fbf01781 100644 (file)
@@ -2375,7 +2375,7 @@ static void perf_event_context_sched_out(struct task_struct *task, int ctxn,
        next_parent = rcu_dereference(next_ctx->parent_ctx);
 
        /* If neither context have a parent context; they cannot be clones. */
-       if (!parent || !next_parent)
+       if (!parent && !next_parent)
                goto unlock;
 
        if (next_parent == ctx || next_ctx == parent || next_parent == parent) {