Move Task Scheduler Single Thread Task Runners to Dedicated Threads

base/task_scheduler/task_scheduler_impl_unittest.cc

Index: base/task_scheduler/task_scheduler_impl_unittest.cc
diff --git a/base/task_scheduler/task_scheduler_impl_unittest.cc b/base/task_scheduler/task_scheduler_impl_unittest.cc
index cda06063ed7444fee2a4d7b2c4818332095f7313..201dcb7c42f1dee39fc4cd94411fe7be415369e7 100644
--- a/base/task_scheduler/task_scheduler_impl_unittest.cc
+++ b/base/task_scheduler/task_scheduler_impl_unittest.cc
@@ -277,6 +277,13 @@ TEST_P(TaskSchedulerImplTest, PostTasksViaTaskRunner) {
   }
 
   factory.WaitForAllTasksToRun();
+
+  // State like Single Thread Task Runners may be lingering around at this point
+  // attached to tasks on their running thread even though
+  // WaitForAllTasksToRun() returned. This will cause issues when Single Thread
+  // Task Runner unregistration occurs during JoinForTesting(). Flushing out the
+  // tasks prevents this state.
+  scheduler_->FlushForTesting();

[fdoray, 2017/01/27 16:47:35]

Is the issue that this DCHECK can be hit
https://codereview.chromium.org/2650383007/diff/60001/base/task_scheduler/sch...
if a SingleThreadTaskRunner is destroyed during
SchedulerSingleThreadWorkerPoolManager::JoinForTesting()?

If so, could you instead add a |join_for_testing_called_| flag in
SchedulerSingleThreadWorkerPoolManager and return immediately in
UnregisterSingleThreadWorkerPool() when this flag is set? That way, we won't
have to think about this issue every time we join TaskScheduler.

[robliao, 2017/01/27 21:25:41]

sgtm. Done. Added a reinforcing comment in task_scheduler.h to say that after
JoinForTesting() returns, that's the end for that task scheduler instance.

 }
 
 INSTANTIATE_TEST_CASE_P(OneTraitsExecutionModePair,