Remove CompositeAndReadback from LayerTreeHost(Impl) and the Proxys.

Remove CompositeAndReadback from LayerTreeHost(Impl) and the Proxys.

R=brianderson, enne
BUG=251936
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=270841

[danakj, 2014-05-09 23:18:10]

This will be step 1 to remove C&R. It doesn't touch the scheduler or the
interfaces between the scheduler and the proxy, but after this the scheduler
will have no way to get into C&R states outside of unit tests.

[danakj, 2014-05-09 23:22:10]

This is not landable until android CompositorImpl stops calling
CompositeAndReadback. Will figure out how to make sure that happens
post-branch-point.

[brianderson, 2014-05-10 00:47:53]

Didn't get to look at it all yet, but this is awesome. Thanks Dana.

https://codereview.chromium.org/279013002/diff/20001/cc/output/direct_renderer.h
File cc/output/direct_renderer.h (left):

https://codereview.chromium.org/279013002/diff/20001/cc/output/direct_rendere...
cc/output/direct_renderer.h:29: virtual bool CanReadPixels() const OVERRIDE;
Does the DirectRenderer not have a GetFramebufferPixels or equivalent to remove?

https://codereview.chromium.org/279013002/diff/20001/cc/output/gl_renderer.cc
File cc/output/gl_renderer.cc (left):

https://codereview.chromium.org/279013002/diff/20001/cc/output/gl_renderer.cc...
cc/output/gl_renderer.cc:1961: COMPILE_ASSERT(sizeof(Float4) == 4 *
sizeof(float),  // NOLINT(runtime/sizeof)
Unrelated changes?

https://codereview.chromium.org/279013002/diff/20001/cc/trees/layer_tree_host...
File cc/trees/layer_tree_host_impl.h (right):

https://codereview.chromium.org/279013002/diff/20001/cc/trees/layer_tree_host...
cc/trees/layer_tree_host_impl.h:195: virtual DrawSwapReadbackResult::DrawResult
PrepareToDraw(FrameData* frame);
I would say rename this to DrawSwapResult, but Simon is in the process of
removing this already.

[danakj, 2014-05-10 15:30:54]

https://codereview.chromium.org/279013002/diff/20001/cc/output/direct_renderer.h
File cc/output/direct_renderer.h (left):

https://codereview.chromium.org/279013002/diff/20001/cc/output/direct_rendere...
cc/output/direct_renderer.h:29: virtual bool CanReadPixels() const OVERRIDE;
On 2014/05/10 00:47:53, brianderson wrote:
> Does the DirectRenderer not have a GetFramebufferPixels or equivalent to
remove?

No the call went right to the gl/software renderers since it wasn't a stateful
call just grabbed the back buffer.

https://codereview.chromium.org/279013002/diff/20001/cc/output/gl_renderer.cc
File cc/output/gl_renderer.cc (left):

https://codereview.chromium.org/279013002/diff/20001/cc/output/gl_renderer.cc...
cc/output/gl_renderer.cc:1961: COMPILE_ASSERT(sizeof(Float4) == 4 *
sizeof(float),  // NOLINT(runtime/sizeof)
On 2014/05/10 00:47:53, brianderson wrote:
> Unrelated changes?

Yeh.. The linter complained that size of wasn't a valid linter exception
anymore.

https://codereview.chromium.org/279013002/diff/20001/cc/trees/layer_tree_host...
File cc/trees/layer_tree_host_impl.h (right):

https://codereview.chromium.org/279013002/diff/20001/cc/trees/layer_tree_host...
cc/trees/layer_tree_host_impl.h:195: virtual DrawSwapReadbackResult::DrawResult
PrepareToDraw(FrameData* frame);
On 2014/05/10 00:47:53, brianderson wrote:
> I would say rename this to DrawSwapResult, but Simon is in the process of
> removing this already.

I was planning to change this to a bool or rename it etc in next CL since this
would require changes in the scheduler.

[danakj, 2014-05-12 17:44:21]

I've rebased this over the change from DrawSwapReadbackResult to DrawSwapResult

[enne (OOO), 2014-05-12 18:21:50]

lgtm

[brianderson, 2014-05-12 20:41:40]

lgtm

[danakj, 2014-05-15 19:37:12]

The CQ bit was checked by danakj@chromium.org

[commit-bot: I haz the power, 2014-05-15 19:38:23]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/danakj@chromium.org/279013002/40001

[commit-bot: I haz the power, 2014-05-15 23:10:54]

Change committed as 270841