Android: Add support for CompositorFrame swaps

Android: Add support for CompositorFrame swaps

Needs https://codereview.chromium.org/12371002

NOTRY=True

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=186278

[no sievers, 2013-03-01 21:44:32]

Hey Sami, this wires up composite-to-mailbox for Android.

We are probably still sending the ACK a bit early (should probably do it after
commit instead of rightaway) which might lead to the renderer being not
throttled correctly. 

But ideally I want to use the mailbox support in cc::Texture Layer, which would
allow us to elegantly send the ACK after commit (but that breaks the thumbnailer
since we cannot access the texture anymore then, so if that cannot be solved
easily I will have to wire it up to the compositor's didCommit callback
instead).

[Sami, 2013-03-05 17:44:42]

Sorry for the delay. Otherwise looks fine but I wanted to ask about the
closures.

https://codereview.chromium.org/12389061/diff/3001/content/browser/renderer_h...
File content/browser/renderer_host/image_transport_factory_android.cc (right):

https://codereview.chromium.org/12389061/diff/3001/content/browser/renderer_h...
content/browser/renderer_host/image_transport_factory_android.cc:110: if
(!context_->makeContextCurrent()) {
Tempted to suggest something like makeHelperContextCurrent() to avoid a 5th and
6th copy of the same code :) It wouldn't be that much less code so this way is
fine too.

https://codereview.chromium.org/12389061/diff/3001/content/browser/renderer_h...
File content/browser/renderer_host/render_widget_host_view_android.cc (right):

https://codereview.chromium.org/12389061/diff/3001/content/browser/renderer_h...
content/browser/renderer_host/render_widget_host_view_android.cc:57: const
gfx::Size& return_size) {
Nit: size by value.

https://codereview.chromium.org/12389061/diff/3001/content/browser/renderer_h...
content/browser/renderer_host/render_widget_host_view_android.cc:496: 
Maybe I'm missing something but why not call
InsertSyncPointAndAckForCompositor() here directly instead of packaging it into
a closure?

https://codereview.chromium.org/12389061/diff/3001/content/browser/renderer_h...
content/browser/renderer_host/render_widget_host_view_android.cc:521:
BuffersSwapped(mailbox, params.size, callback);
Same Q about closure usage.

https://codereview.chromium.org/12389061/diff/3001/content/browser/renderer_h...
File content/browser/renderer_host/render_widget_host_view_android.h (right):

https://codereview.chromium.org/12389061/diff/3001/content/browser/renderer_h...
content/browser/renderer_host/render_widget_host_view_android.h:180: const
gfx::Size& size,
Nit: pass size by value.

[Sami, 2013-03-05 19:19:10]

Had an offline clarification that the closure is for future expansion where we
can have the ack happen later like it should. This patch lgmt modulo the above
nits.

[Sami, 2013-03-05 19:19:22]

Erm, lgtm.

[commit-bot: I haz the power, 2013-03-05 21:57:53]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/sievers@chromium.org/12389061/11001

[no sievers, 2013-03-05 21:58:05]

https://codereview.chromium.org/12389061/diff/3001/content/browser/renderer_h...
File content/browser/renderer_host/image_transport_factory_android.cc (right):

https://codereview.chromium.org/12389061/diff/3001/content/browser/renderer_h...
content/browser/renderer_host/image_transport_factory_android.cc:110: if
(!context_->makeContextCurrent()) {
On 2013/03/05 17:44:42, Sami wrote:
> Tempted to suggest something like makeHelperContextCurrent() to avoid a 5th
and
> 6th copy of the same code :) It wouldn't be that much less code so this way is
> fine too.

I wish I could eliminate the helper context...

https://codereview.chromium.org/12389061/diff/3001/content/browser/renderer_h...
File content/browser/renderer_host/render_widget_host_view_android.cc (right):

https://codereview.chromium.org/12389061/diff/3001/content/browser/renderer_h...
content/browser/renderer_host/render_widget_host_view_android.cc:57: const
gfx::Size& return_size) {
On 2013/03/05 17:44:42, Sami wrote:
> Nit: size by value.

Done.

https://codereview.chromium.org/12389061/diff/3001/content/browser/renderer_h...
File content/browser/renderer_host/render_widget_host_view_android.h (right):

https://codereview.chromium.org/12389061/diff/3001/content/browser/renderer_h...
content/browser/renderer_host/render_widget_host_view_android.h:180: const
gfx::Size& size,
On 2013/03/05 17:44:42, Sami wrote:
> Nit: pass size by value.

Done.

[commit-bot: I haz the power, 2013-03-05 22:55:57]

Retried try job too often on mac_rel for step(s) interactive_ui_tests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=mac_rel&nu...

[commit-bot: I haz the power, 2013-03-05 23:03:29]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/sievers@chromium.org/12389061/11001

[commit-bot: I haz the power, 2013-03-05 23:10:40]

Change committed as 186278