perf: Fix NULL deref

Dan reported:

  1229                  if (ctx->task == TASK_TOMBSTONE ||
  1230                      !atomic_inc_not_zero(&ctx->refcount)) {
  1231                          raw_spin_unlock(&ctx->lock);
  1232                          ctx = NULL;
                                ^^^^^^^^^^
ctx is NULL.

  1233                  }
  1234
  1235                  WARN_ON_ONCE(ctx->task != task);
                                     ^^^^^^^^^^^^^^^^^
The patch adds a NULL dereference.

Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: David Ahern <dsahern@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Stephane Eranian <eranian@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Vince Weaver <vincent.weaver@maine.edu>
Fixes: 63b6da3 ("perf: Fix perf_event_exit_task() race")
Signed-off-by: Ingo Molnar <mingo@kernel.org>

kernel/events/core.c

@@ -1230,9 +1230,9 @@ perf_lock_task_context(struct task_struct *task, int ctxn, unsigned long *flags)
 		    !atomic_inc_not_zero(&ctx->refcount)) {
 			raw_spin_unlock(&ctx->lock);
 			ctx = NULL;
+		} else {
+			WARN_ON_ONCE(ctx->task != task);
 		}
-
-		WARN_ON_ONCE(ctx->task != task);
 	}
 	rcu_read_unlock();
 	if (!ctx)