cgroup: fix rcu_read_lock() leak in update_if_frozen()

While updating cgroup_freezer locking, 68fafb77d827 ("cgroup_freezer:
replace freezer->lock with freezer_mutex") introduced a bug in
update_if_frozen() where it returns with rcu_read_lock() held.  Fix it
by adding rcu_read_unlock() before returning.
Signed-off-by: Tejun Heo <tj@kernel.org>
Reported-by: kbuild test robot <fengguang.wu@intel.com>

kernel/cgroup_freezer.c

@@ -267,9 +267,11 @@ static void update_if_frozen(struct cgroup_subsys_state *css)
 		struct freezer *child = css_freezer(pos);
 
 		if ((child->state & CGROUP_FREEZER_ONLINE) &&
-		    !(child->state & CGROUP_FROZEN))
+		    !(child->state & CGROUP_FROZEN)) {
+			rcu_read_unlock();
 			return;
 		}
+	}
 	rcu_read_unlock();
 
 	/* are all tasks frozen? */