Update AudioRenderer, VideoRenderer, and AudioDecoder Initialize() methods to use PipelineStatusCB.

Update AudioRenderer, VideoRenderer, and AudioDecoder Initialize() methods to use PipelineStatusCB.

- Fixes race on state_ in CompositeFilter detected by TSAN
- Moves us one step closer to removing FilterHost::SetError().

BUG=109875
TEST=AudioRendererBaseTest.* VideoRendererBaseTest.*, FFmpegAudioDecoderTest.*, CompositeFilterTest.*


Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=120135

[Ami GONE FROM CHROMIUM, 2012-02-01 18:39:11]

LGTM  \o/

http://codereview.chromium.org/9310028/diff/1/media/base/composite_filter.cc
File media/base/composite_filter.cc (right):

http://codereview.chromium.org/9310028/diff/1/media/base/composite_filter.cc#...
media/base/composite_filter.cc:452: DCHECK_EQ(message_loop_,
MessageLoop::current());
DCHECK_NE(state_, kCreated);

http://codereview.chromium.org/9310028/diff/1/media/base/mock_filters.cc
File media/base/mock_filters.cc (right):

http://codereview.chromium.org/9310028/diff/1/media/base/mock_filters.cc#newc...
media/base/mock_filters.cc:176: callback.Run(PIPELINE_OK);
The fact that this method doesn't take a |status| param (and always fires
PIPELINE_OK) is suspicious (might simply indicate a lack of test coverage).

http://codereview.chromium.org/9310028/diff/1/media/base/pipeline_unittest.cc
File media/base/pipeline_unittest.cc (right):

http://codereview.chromium.org/9310028/diff/1/media/base/pipeline_unittest.cc...
media/base/pipeline_unittest.cc:157: .WillOnce(Invoke(&RunPipelineStatusCB3));
Is the fact that this worked before not a bug in the callback machinery?

[Ami GONE FROM CHROMIUM, 2012-02-01 18:51:13]

http://codereview.chromium.org/9310028/diff/1/media/base/pipeline_unittest.cc
File media/base/pipeline_unittest.cc (right):

http://codereview.chromium.org/9310028/diff/1/media/base/pipeline_unittest.cc...
media/base/pipeline_unittest.cc:157: .WillOnce(Invoke(&RunPipelineStatusCB3));
On 2012/02/01 18:39:11, Ami Fischman wrote:
> Is the fact that this worked before not a bug in the callback machinery?

nm, I thought this was video *decoder*, not *renderer*.

[acolwell GONE FROM CHROMIUM, 2012-02-01 19:59:46]

http://codereview.chromium.org/9310028/diff/1/media/base/composite_filter.cc
File media/base/composite_filter.cc (right):

http://codereview.chromium.org/9310028/diff/1/media/base/composite_filter.cc#...
media/base/composite_filter.cc:452: DCHECK_EQ(message_loop_,
MessageLoop::current());
On 2012/02/01 18:39:11, Ami Fischman wrote:
> DCHECK_NE(state_, kCreated);

Done.

http://codereview.chromium.org/9310028/diff/1/media/base/mock_filters.cc
File media/base/mock_filters.cc (right):

http://codereview.chromium.org/9310028/diff/1/media/base/mock_filters.cc#newc...
media/base/mock_filters.cc:176: callback.Run(PIPELINE_OK);
On 2012/02/01 18:39:11, Ami Fischman wrote:
> The fact that this method doesn't take a |status| param (and always fires
> PIPELINE_OK) is suspicious (might simply indicate a lack of test coverage).
Agreed. Created a bug for adding initialization failure tests.
http://crbug.com/112337

[commit-bot: I haz the power, 2012-02-01 21:50:55]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/acolwell@chromium.org/9310028/5002

[commit-bot: I haz the power, 2012-02-02 00:37:43]

Change committed as 120135