[SPARK-5069][Core] Fix the race condition of TaskSchedulerImpl.dagScheduler

It's not necessary to set `TaskSchedulerImpl.dagScheduler` in preStart. It's safe to set it after `initializeEventProcessActor()`.

Author: zsxwing <zsxwing@gmail.com>

Closes #3887 from zsxwing/SPARK-5069 and squashes the following commits:

d95894f [zsxwing] Fix the race condition of TaskSchedulerImpl.dagScheduler

core/src/main/scala/org/apache/spark/scheduler/DAGScheduler.scala

@@ -138,6 +138,7 @@ class DAGScheduler(
   }
 
   initializeEventProcessActor()
+  taskScheduler.setDAGScheduler(this)
 
   // Called by TaskScheduler to report task's starting.
   def taskStarted(task: Task[_], taskInfo: TaskInfo) {
@@ -1375,12 +1376,6 @@ private[scheduler] class DAGSchedulerActorSupervisor(dagScheduler: DAGScheduler)
 private[scheduler] class DAGSchedulerEventProcessActor(dagScheduler: DAGScheduler)
   extends Actor with Logging {
 
-  override def preStart() {
-    // set DAGScheduler for taskScheduler to ensure eventProcessActor is always
-    // valid when the messages arrive
-    dagScheduler.taskScheduler.setDAGScheduler(dagScheduler)
-  }
-
   /**
    * The main event loop of the DAG scheduler.
    */

core/src/test/scala/org/apache/spark/scheduler/TaskSchedulerImplSuite.scala

@@ -305,7 +305,6 @@ class TaskSchedulerImplSuite extends FunSuite with LocalSparkContext with Loggin
       override def taskStarted(task: Task[_], taskInfo: TaskInfo) {}
       override def executorAdded(execId: String, host: String) {}
     }
-    taskScheduler.setDAGScheduler(dagScheduler)
     // Give zero core offers. Should not generate any tasks
     val zeroCoreWorkerOffers = Seq(new WorkerOffer("executor0", "host0", 0),
       new WorkerOffer("executor1", "host1", 0))