Don't set scroll client for main frame

Don't set scroll client for main frame

Instead, only set the scroll client on scrollable layers that are not
the main frame.  This also drops a needless per-layer allocation for
non-scrollable layers.

By doing this, it allows LayerTreeHost to reuse
cc::Layer::SetScrollOffsetFromImpl for pre-setting the value of scroll
offset on the root scroll layer prior to calling ApplyScrollAndScale.

This patch depends on https://codereview.chromium.org/18348021/.

R=jamesr@chromium.org
BUG=256381
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=153698

[jamesr, 2013-07-04 00:37:36]

https://codereview.chromium.org/18083034/diff/1/Source/core/page/scrolling/Sc...
File Source/core/page/scrolling/ScrollingCoordinator.cpp (right):

https://codereview.chromium.org/18083034/diff/1/Source/core/page/scrolling/Sc...
Source/core/page/scrolling/ScrollingCoordinator.cpp:540: return scrollableArea
== static_cast<ScrollableArea*>(m_page->mainFrame()->view());
do you need the static cast on the RHS of this for the comparison to compile?
the compiler knows the types on each side so it should be able to do the
mangling necessary (which should only be offset adjustments in case of multiple
inheritance, which we don't have here).

https://codereview.chromium.org/18083034/diff/1/Source/core/platform/graphics...
File Source/core/platform/graphics/GraphicsLayer.cpp (right):

https://codereview.chromium.org/18083034/diff/1/Source/core/platform/graphics...
Source/core/platform/graphics/GraphicsLayer.cpp:1348:
m_layer->layer()->setScrollClient(nullptr);
setScrollClient() just takes a WebLayerScrollClient*. I'm not super confident
that nullptr can be used for this on all of our toolchains, although we want to
get there eventually. keep a close eye on it after landing or just use '0' for
now

[jamesr, 2013-07-04 00:37:42]

lgtm

[enne (OOO), 2013-07-04 00:50:17]

https://codereview.chromium.org/18083034/diff/1/Source/core/page/scrolling/Sc...
File Source/core/page/scrolling/ScrollingCoordinator.cpp (right):

https://codereview.chromium.org/18083034/diff/1/Source/core/page/scrolling/Sc...
Source/core/page/scrolling/ScrollingCoordinator.cpp:540: return scrollableArea
== static_cast<ScrollableArea*>(m_page->mainFrame()->view());
On 2013/07/04 00:37:37, jamesr wrote:
> do you need the static cast on the RHS of this for the comparison to compile?
> the compiler knows the types on each side so it should be able to do the
> mangling necessary (which should only be offset adjustments in case of
multiple
> inheritance, which we don't have here).

You're right.  The compiler will do the right thing here.  Done.

https://codereview.chromium.org/18083034/diff/1/Source/core/platform/graphics...
File Source/core/platform/graphics/GraphicsLayer.cpp (right):

https://codereview.chromium.org/18083034/diff/1/Source/core/platform/graphics...
Source/core/platform/graphics/GraphicsLayer.cpp:1348:
m_layer->layer()->setScrollClient(nullptr);
On 2013/07/04 00:37:37, jamesr wrote:
> setScrollClient() just takes a WebLayerScrollClient*. I'm not super confident
> that nullptr can be used for this on all of our toolchains, although we want
to
> get there eventually. keep a close eye on it after landing or just use '0' for
> now

Changed to 0.  :(

[commit-bot: I haz the power, 2013-07-08 17:30:23]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/enne@chromium.org/18083034/7001

[commit-bot: I haz the power, 2013-07-08 18:53:37]

Change committed as 153698