Remove renderer-side tracking of accelerated compositing

Remove renderer-side tracking of accelerated compositing

Remove didActivate/didDeactivateCompositor from everywhere that calls it
as it now does nothing.

BUG=none

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=271234

[enne (OOO), 2014-05-14 00:21:30]

I'm not totally sure what I'm doing here, but it doesn't seem like any of this
code should exist anymore.  Layout tests are the only thing that aren't force
composited yet and they don't use these RenderWidget code paths

I'm happy to clean up the Blink calling code first to make sure this never gets
called if that makes this patch more comfortable to land, but that would need to
wait on layout tests.

[piman, 2014-05-14 00:30:00]

On 2014/05/14 00:21:30, enne wrote:
> I'm not totally sure what I'm doing here, but it doesn't seem like any of this
> code should exist anymore.  Layout tests are the only thing that aren't force
> composited yet and they don't use these RenderWidget code paths
> 
> I'm happy to clean up the Blink calling code first to make sure this never
gets
> called if that makes this patch more comfortable to land, but that would need
to
> wait on layout tests.

I already have https://codereview.chromium.org/274163004/ for the browser side
of this (including removing IPCs).

It has a couple of test failures though, that I was looking at before landing
(racy tests).

[danakj, 2014-05-14 15:03:29]

On 2014/05/14 00:30:00, piman wrote:
> On 2014/05/14 00:21:30, enne wrote:
> > I'm not totally sure what I'm doing here, but it doesn't seem like any of
this
> > code should exist anymore.  Layout tests are the only thing that aren't
force
> > composited yet and they don't use these RenderWidget code paths

LayoutTests do call didActivate/didDeactivateCompositor from WebViewImpl, but
the important stuff (SetDeferredCommits) is done directly from WebViewImpl, so
making these changes looks good to me.

[danakj, 2014-05-14 15:03:34]

https://codereview.chromium.org/288633002/diff/1/content/renderer/render_widg...
File content/renderer/render_widget_fullscreen_pepper.cc (right):

https://codereview.chromium.org/288633002/diff/1/content/renderer/render_widg...
content/renderer/render_widget_fullscreen_pepper.cc:327: if (!layer_)
Should we setRootLayer(NULL) if !layer_? Before we would have told the browser
we left compositing which would do things like say "there's no surface to copy
from" for a readback and stuff, now it might continue showing the old content
instead of nothing?

https://codereview.chromium.org/288633002/diff/1/content/renderer/render_widg...
content/renderer/render_widget_fullscreen_pepper.cc:331: if (!layerTreeView())
Can this go away too?

[enne (OOO), 2014-05-15 20:51:14]

Rebased on top of piman's patch.

https://codereview.chromium.org/288633002/diff/1/content/renderer/render_widg...
File content/renderer/render_widget_fullscreen_pepper.cc (right):

https://codereview.chromium.org/288633002/diff/1/content/renderer/render_widg...
content/renderer/render_widget_fullscreen_pepper.cc:327: if (!layer_)
On 2014/05/14 15:03:34, danakj wrote:
> Should we setRootLayer(NULL) if !layer_? Before we would have told the browser
> we left compositing which would do things like say "there's no surface to copy
> from" for a readback and stuff, now it might continue showing the old content
> instead of nothing?

Sure.  I think that seems reasonable, although I think it didn't do that before.

https://codereview.chromium.org/288633002/diff/1/content/renderer/render_widg...
content/renderer/render_widget_fullscreen_pepper.cc:331: if (!layerTreeView())
On 2014/05/14 15:03:34, danakj wrote:
> Can this go away too?

Done.

[piman, 2014-05-15 21:16:03]

lgtm

[danakj, 2014-05-15 21:18:07]

LGTM2

[enne (OOO), 2014-05-16 22:56:24]

The CQ bit was checked by enne@chromium.org

[commit-bot: I haz the power, 2014-05-16 23:33:46]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/enne@chromium.org/288633002/20001

[commit-bot: I haz the power, 2014-05-17 17:08:06]

Change committed as 271234