Add TaskRunner wrapper for WorkerPool.

Add TaskRunner wrapper for WorkerPool.

BUG=114329
TEST=none


Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=133855

[willchan no longer on Chromium, 2012-04-24 23:41:50]

http://codereview.chromium.org/10220001/diff/5003/base/threading/worker_pool.cc
File base/threading/worker_pool.cc (right):

http://codereview.chromium.org/10220001/diff/5003/base/threading/worker_pool....
base/threading/worker_pool.cc:54: bool PostDelayedTaskAssertZeroDelay(
Can you kill off one of these? You can do the conversion in the virtual member
function.

http://codereview.chromium.org/10220001/diff/5003/base/threading/worker_pool.h
File base/threading/worker_pool.h (right):

http://codereview.chromium.org/10220001/diff/5003/base/threading/worker_pool....
base/threading/worker_pool.h:56: static scoped_refptr<TaskRunner>
GetTaskRunner(bool task_is_slow);
How about you also store a leaky LazyInstance for a scoped_refptr for the
TaskRunner? I worry that people will do dumb things like TaskRunner* task_runner
= base::WorkerPool::GetTaskRunner(), which will cause the TaskRunner to get
immediately deleted due to the implicit T* operator.

[mattm, 2012-04-25 00:57:35]

http://codereview.chromium.org/10220001/diff/5003/base/threading/worker_pool.cc
File base/threading/worker_pool.cc (right):

http://codereview.chromium.org/10220001/diff/5003/base/threading/worker_pool....
base/threading/worker_pool.cc:54: bool PostDelayedTaskAssertZeroDelay(
On 2012/04/24 23:41:50, willchan wrote:
> Can you kill off one of these? You can do the conversion in the virtual member
> function.

Done.

http://codereview.chromium.org/10220001/diff/5003/base/threading/worker_pool.h
File base/threading/worker_pool.h (right):

http://codereview.chromium.org/10220001/diff/5003/base/threading/worker_pool....
base/threading/worker_pool.h:56: static scoped_refptr<TaskRunner>
GetTaskRunner(bool task_is_slow);
On 2012/04/24 23:41:50, willchan wrote:
> How about you also store a leaky LazyInstance for a scoped_refptr for the
> TaskRunner? I worry that people will do dumb things like TaskRunner*
task_runner
> = base::WorkerPool::GetTaskRunner(), which will cause the TaskRunner to get
> immediately deleted due to the implicit T* operator.

Done.  Had to make a holder class to initialize the scoped_refptr

[willchan no longer on Chromium, 2012-04-25 01:01:08]

lgtm

http://codereview.chromium.org/10220001/diff/14001/base/threading/worker_pool.cc
File base/threading/worker_pool.cc (right):

http://codereview.chromium.org/10220001/diff/14001/base/threading/worker_pool...
base/threading/worker_pool.cc:62: TimeDelta delay);
You killed a definition, but not the declaration.

[mattm, 2012-04-25 01:21:25]

http://codereview.chromium.org/10220001/diff/14001/base/threading/worker_pool.cc
File base/threading/worker_pool.cc (right):

http://codereview.chromium.org/10220001/diff/14001/base/threading/worker_pool...
base/threading/worker_pool.cc:62: TimeDelta delay);
On 2012/04/25 01:01:08, willchan wrote:
> You killed a definition, but not the declaration.

oops, done.

[commit-bot: I haz the power, 2012-04-25 01:21:37]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/mattm@chromium.org/10220001/4009

[commit-bot: I haz the power, 2012-04-25 03:15:08]

Change committed as 133855