Move ownership of cc:VideoLayer to webkit/media

Move ownership of cc:VideoLayer to webkit/media

Currently the cc::VideoLayer is created and owned by the WebMediaPlayerClientImpl
in WebKit. Communication with the cc::VideoLayerImpl then must be done though the
WebVideoFrame and WebStreamTextureClient APIs.

This CL moves the creation and ownership of the cc::Video layer to the
WebMediaPlayer implementations in webkit/media, but does so behind a compiler
flag until the WebKit side to match it lands and enables the flag.

Then the communication between the WebMediaPlayer and cc::VideoLayerImpl can be
done directly through the cc::VideoFrameProvider interface. This interface is
implemented in the various implementations of WebMediaPlayer behind the
compiler flag as well, to be enabled when WebMediaPlayerClientImpl stops
implementing cc::VideoFrameProvider via WebVideoFrameProvider.

Once the WebKit side lands and rolls, we can completely remove all mention of
WebVideoFrame, WebVideoFrameProvider, and WebStreamTextureClientImpl.

R=jamesr
BUG=196810
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=189979

[danakj, 2013-03-21 21:36:55]

This is just the bahaviour pieces of https://codereview.chromium.org/12902002/
behind flags. New code is ifdef'd. Old code is ifndef'd.

After webkit side turns the flags on, we can remove all the stuff in ifndefs, as
well as all the classes supporting WebVideoFrame(Provider).

[jamesr, 2013-03-22 00:11:13]

lgtm!  Be sure to carefully check that every file here that tests the #define
includes the header that you want to set the #define from.

https://codereview.chromium.org/12676004/diff/1010/webkit/media/webmediaplaye...
File webkit/media/webmediaplayer_ms.cc (right):

https://codereview.chromium.org/12676004/diff/1010/webkit/media/webmediaplaye...
webkit/media/webmediaplayer_ms.cc:404: DVLOG(3) <<
"WebMediaPlayerMS::getCurrentFrame";
GetCurrentFrame ?

https://codereview.chromium.org/12676004/diff/1010/webkit/media/webmediaplaye...
webkit/media/webmediaplayer_ms.cc:416: DVLOG(3) <<
"WebMediaPlayerMS::putCurrentFrame";
PutCurrentFrame?

[danakj, 2013-03-22 04:12:27]

On 2013/03/22 00:11:13, jamesr wrote:
> lgtm!  Be sure to carefully check that every file here that tests the #define
> includes the header that you want to set the #define from.

I added WebVideoFrame.h to webkit/media/android/stream_texture_factory_android.h
for now, so that it's everywhere.

[danakj, 2013-03-22 04:12:31]

https://codereview.chromium.org/12676004/diff/1010/webkit/media/webmediaplaye...
File webkit/media/webmediaplayer_ms.cc (right):

https://codereview.chromium.org/12676004/diff/1010/webkit/media/webmediaplaye...
webkit/media/webmediaplayer_ms.cc:404: DVLOG(3) <<
"WebMediaPlayerMS::getCurrentFrame";
On 2013/03/22 00:11:13, jamesr wrote:
> GetCurrentFrame ?

Done.

https://codereview.chromium.org/12676004/diff/1010/webkit/media/webmediaplaye...
webkit/media/webmediaplayer_ms.cc:416: DVLOG(3) <<
"WebMediaPlayerMS::putCurrentFrame";
On 2013/03/22 00:11:13, jamesr wrote:
> PutCurrentFrame?

Done.

[commit-bot: I haz the power, 2013-03-22 04:15:53]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/danakj@chromium.org/12676004/2002

[commit-bot: I haz the power, 2013-03-22 16:47:08]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/danakj@chromium.org/12676004/2002

[commit-bot: I haz the power, 2013-03-22 18:32:04]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/danakj@chromium.org/12676004/2002

[danakj, 2013-03-23 03:30:01]

Committed patchset #5 manually as r189979 (presubmit successful).