Make new TaskRunner, SequencedTaskRunner, and SingleThreadTaskRunner interfaces

base/task_runner.h

Index: base/task_runner.h
diff --git a/base/task_runner.h b/base/task_runner.h
new file mode 100644
index 0000000000000000000000000000000000000000..0b3e0cb6d344b1d0923a59a77f9f8917f26911f1
--- /dev/null
+++ b/base/task_runner.h
@@ -0,0 +1,160 @@
+// Copyright (c) 2012 The Chromium Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file.
+
+#ifndef BASE_TASK_RUNNER_H_
+#define BASE_TASK_RUNNER_H_
+#pragma once
+
+#include "base/base_export.h"
+#include "base/basictypes.h"
+#include "base/callback_forward.h"
+#include "base/memory/ref_counted.h"
+
+namespace tracked_objects {
+class Location;
+} // namespace tracked_objects
+
+namespace base {
+
+struct TaskRunnerTraits;
+
+// A TaskRunner is an object that runs submitted tasks (in the form of

[willchan no longer on Chromium, 2012/02/09 20:50:50]

s/submitted/posted/

[akalin, 2012/02/10 22:48:59]

Done.

+// Closure objects).  The TaskRunner interface provides a way of
+// decoupling task submission from the mechanics of how each task will
+// be run.  TaskRunner provides very weak guarantees as to how
+// submitted tasks are run (or if they're run at all).  In particular,
+// it only guarantees:
+//
+//   - Submitting a task will not run it synchronously.  That is, no

[willchan no longer on Chromium, 2012/02/09 20:50:50]

Replace all the submit with post.

[akalin, 2012/02/10 22:48:59]

Done.

+//     Post*Task method will call task.Run() directly.
+//
+//   - Increasing the delay can only delay when the task gets run.

[willchan no longer on Chromium, 2012/02/09 20:50:50]

It seems like this section is unnecessary (too obvious). Is there a subtlety I'm
not noticing?

"it won't make it run earlier than it normally would"?

[akalin, 2012/02/10 22:48:59]

It is pretty obvious, but I think it's worthwhile to state.  It implies that
implementations may either ignore the parameter or use it, but only in such a
way that it actually works.

+//     That is, increasing the delay may not affect when the task gets
+//     run, or it could make it run later than it normally would, but
+//     it won't make it run earlier than it normally would.
+//
+// TaskRunner does not guarantee the order in which submitted tasks
+// are run, whether tasks overlap, or whether they're run on a
+// particular thread.  Also it does not guarantee a memory model for
+// shared data between tasks.  (In other words, you should use your
+// own synchronization/locking primitives if you need to share data
+// between tasks.)
+//
+// Implementations of TaskRunner should be thread-safe in that all
+// methods must be safe to call on any thread.  Ownership semantics
+// for TaskRunners are in general not clear, which is why the
+// interface itself is RefCountedThreadSafe.
+//
+// Some theoretical implementations of TaskRunner:
+//
+//   - A TaskRunner that uses a worker pool to run submitted tasks.

[willchan no longer on Chromium, 2012/02/09 20:50:50]

s/worker pool/thread pool/ to make it more obvious it can be multithreaded.

[akalin, 2012/02/10 22:48:59]

Done.

+//
+//   - A TaskRunner that, for each task, spawns a non-joinable thread
+//     to run that task and immediately quit.
+//
+//   - A TaskRunner that stores the list of submitted tasks and has a
+//     method Run() that runs each runnable task in random order.
+class BASE_EXPORT TaskRunner
+    : public RefCountedThreadSafe<TaskRunner, TaskRunnerTraits> {
+ public:
+  // TODO(akalin): Get rid of the boolean return value for the
+  // Post*Task methods.
+
+  // Submits the given task for execution.  Returns true if the task
+  // may be run at some point in the future, and false if the task
+  // definitely will not be run.
+  //
+  // TODO(akalin): Consider forwarding this to PostDelayedTask(_, _,

[willchan no longer on Chromium, 2012/02/09 20:50:50]

What's your thought here? Do it or not? We should delete the TODO and just do
one of these. I kinda like eliminating the virtual.

[akalin, 2012/02/10 22:48:59]

Okay, I just implemented this TODO.

+  // 0).
+  virtual bool PostTask(const tracked_objects::Location& from_here,
+                        const Closure& task) = 0;
+
+  // Like PostTask, but tries to run the submitted task only after
+  // |delay_ms| has passed.  PostDelayedTask with zero delay should be
+  // equivalent to PostTask.
+  //
+  // It is valid for an implementation to ignore |delay_ms|; that is,
+  // to have PostDelayedTask behave the same as PostTask.
+  //
+  // TODO(akalin): Make PostDelayedTask use TimeDelta instead.
+  virtual bool PostDelayedTask(const tracked_objects::Location& from_here,
+                               const Closure& task,
+                               int64 delay_ms) = 0;
+
+  // Returns true if the current thread is a thread on which a task
+  // may be run, and false if no task will be run on the current
+  // thread.
+  //
+  // It is valid for an implementation to always return true, or in
+  // general to use 'true' as a default value.
+  virtual bool RunsTasksOnCurrentThread() const = 0;
+
+  // Posts |task| on the current TaskRunner.  On completion, |reply|
+  // is posted to the thread that called PostTaskAndReply().  Both
+  // |task| and |reply| are guaranteed to be deleted on the thread
+  // from which PostTaskAndReply() is invoked.  This allows objects
+  // that must be deleted on the originating thread to be bound into
+  // the |task| and |reply| Closures.  In particular, it can be useful
+  // to use WeakPtr<> in the |reply| Closure so that the reply
+  // operation can be canceled. See the following pseudo-code:
+  //
+  // class DataBuffer : public RefCountedThreadSafe<DataBuffer> {
+  //  public:
+  //   // Called to add data into a buffer.
+  //   void AddData(void* buf, size_t length);
+  //   ...
+  // };
+  //
+  //
+  // class DataLoader : public SupportsWeakPtr<DataLoader> {
+  //  public:
+  //    void GetData() {
+  //      scoped_refptr<DataBuffer> buffer = new DataBuffer();
+  //      target_thread_.message_loop_proxy()->PostTaskAndReply(
+  //          FROM_HERE,
+  //          base::Bind(&DataBuffer::AddData, buffer),
+  //          base::Bind(&DataLoader::OnDataReceived, AsWeakPtr(), buffer));
+  //    }
+  //
+  //  private:
+  //    void OnDataReceived(scoped_refptr<DataBuffer> buffer) {
+  //      // Do something with buffer.
+  //    }
+  // };
+  //
+  //
+  // Things to notice:
+  //   * Results of |task| are shared with |reply| by binding a shared argument
+  //     (a DataBuffer instance).
+  //   * The DataLoader object has no special thread safety.
+  //   * The DataLoader object can be deleted while |task| is still running,
+  //     and the reply will cancel itself safely because it is bound to a
+  //     WeakPtr<>.
+  bool PostTaskAndReply(const tracked_objects::Location& from_here,
+                        const Closure& task,
+                        const Closure& reply);
+
+ protected:
+  friend struct TaskRunnerTraits;
+
+  // Only the Windows debug build seems to need this: see
+  // http://crbug.com/112250.
+  friend class RefCountedThreadSafe<TaskRunner, TaskRunnerTraits>;
+
+  TaskRunner();
+  virtual ~TaskRunner();
+
+  // Called when this object should be destroyed.  By default simply
+  // deletes |this|, but can be overridden to do something else, like
+  // delete on a certain thread.
+  virtual void OnDestruct() const;
+};
+
+struct BASE_EXPORT TaskRunnerTraits {
+  static void Destruct(const TaskRunner* task_runner);
+};
+
+}  // namespace base
+
+#endif  // BASE_TASK_RUNNER_H_