[Slimming Paint] Track clip renderers

[Slimming Paint] Track clip renderers

This patch updates the clip recorder and clip items to track a
RenderLayerModelObject (a subclass of RenderObject) so invalidations
are properly handled in the update algorithm just like regular
DisplayItems. Clipper callsites have been updated accordingly.

With this patch, the ASSERT in ViewDisplayList::findDisplayItem can
now be removed.

TEST=ViewDisplayListTest_UpdateClip

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=185341

[pdr., 2014-11-13 01:30:19]

pdr@chromium.org changed reviewers:
+ chrishtr@chromium.org, leviw@chromium.org

[chrishtr, 2014-11-13 01:46:11]

https://codereview.chromium.org/719353004/diff/1/Source/core/paint/ClipRecord...
File Source/core/paint/ClipRecorder.cpp (right):

https://codereview.chromium.org/719353004/diff/1/Source/core/paint/ClipRecord...
Source/core/paint/ClipRecorder.cpp:40:
m_renderLayer->renderer()->view()->viewDisplayList().add(adoptPtr(m_clipDisplayItem));
Just use m_renderLayer->renderer() rather than passing the renderer explicitly?

[leviw_travelin_and_unemployed, 2014-11-13 01:46:40]

https://codereview.chromium.org/719353004/diff/1/Source/core/paint/ClipRecord...
File Source/core/paint/ClipRecorder.cpp (right):

https://codereview.chromium.org/719353004/diff/1/Source/core/paint/ClipRecord...
Source/core/paint/ClipRecorder.cpp:40:
m_renderLayer->renderer()->view()->viewDisplayList().add(adoptPtr(m_clipDisplayItem));
Remove m_renderLayer and assert we have a layer.

https://codereview.chromium.org/719353004/diff/1/Source/core/paint/ClipRecord...
File Source/core/paint/ClipRecorder.h (right):

https://codereview.chromium.org/719353004/diff/1/Source/core/paint/ClipRecord...
Source/core/paint/ClipRecorder.h:22: ClipDisplayItem(const RenderObject*
renderer, RenderLayer*, Type type, IntRect clipRect)
This should be a RenderLayerModelObject.

https://codereview.chromium.org/719353004/diff/1/Source/core/paint/ClipRecord...
Source/core/paint/ClipRecorder.h:55: RenderLayer* m_renderLayer;
We can kill the RenderLayer member, no?

[pdr., 2014-11-13 06:49:29]

PTAL

https://codereview.chromium.org/719353004/diff/1/Source/core/paint/ClipRecord...
File Source/core/paint/ClipRecorder.cpp (right):

https://codereview.chromium.org/719353004/diff/1/Source/core/paint/ClipRecord...
Source/core/paint/ClipRecorder.cpp:40:
m_renderLayer->renderer()->view()->viewDisplayList().add(adoptPtr(m_clipDisplayItem));
On 2014/11/13 at 01:46:11, chrishtr wrote:
> Just use m_renderLayer->renderer() rather than passing the renderer
explicitly?

I switched this to not track the render layer at all but instead just track the
renderer directly.

https://codereview.chromium.org/719353004/diff/1/Source/core/paint/ClipRecord...
File Source/core/paint/ClipRecorder.h (right):

https://codereview.chromium.org/719353004/diff/1/Source/core/paint/ClipRecord...
Source/core/paint/ClipRecorder.h:22: ClipDisplayItem(const RenderObject*
renderer, RenderLayer*, Type type, IntRect clipRect)
On 2014/11/13 at 01:46:40, leviw wrote:
> This should be a RenderLayerModelObject.

Done

https://codereview.chromium.org/719353004/diff/1/Source/core/paint/ClipRecord...
Source/core/paint/ClipRecorder.h:55: RenderLayer* m_renderLayer;
On 2014/11/13 at 01:46:40, leviw wrote:
> We can kill the RenderLayer member, no?

R.I.P. RenderLayer member.

[mstensho (USE GERRIT), 2014-11-13 08:32:16]

mstensho@opera.com changed reviewers:
+ mstensho@opera.com

[mstensho (USE GERRIT), 2014-11-13 08:32:16]

https://codereview.chromium.org/719353004/diff/20001/Source/core/paint/LayerP...
File Source/core/paint/LayerPainter.cpp (right):

https://codereview.chromium.org/719353004/diff/20001/Source/core/paint/LayerP...
Source/core/paint/LayerPainter.cpp:104: clipRecorder = adoptPtr(new
ClipRecorder(m_renderLayer.renderer(), context, DisplayItem::ClipLayerParent,
clipRect));
That should be m_renderLayer.parent()->renderer()

[pdr., 2014-11-13 18:40:48]

PTAL

https://codereview.chromium.org/719353004/diff/20001/Source/core/paint/LayerP...
File Source/core/paint/LayerPainter.cpp (right):

https://codereview.chromium.org/719353004/diff/20001/Source/core/paint/LayerP...
Source/core/paint/LayerPainter.cpp:104: clipRecorder = adoptPtr(new
ClipRecorder(m_renderLayer.renderer(), context, DisplayItem::ClipLayerParent,
clipRect));
On 2014/11/13 at 08:32:16, mstensho wrote:
> That should be m_renderLayer.parent()->renderer()
Good catch. Fixed.

[chrishtr, 2014-11-13 18:42:47]

https://codereview.chromium.org/719353004/diff/40001/Source/core/paint/ViewDi...
File Source/core/paint/ViewDisplayListTest.cpp (right):

https://codereview.chromium.org/719353004/diff/40001/Source/core/paint/ViewDi...
Source/core/paint/ViewDisplayListTest.cpp:353: EXPECT_EQ((size_t)4,
firstList.size());
This is tedious and hard to read. Add some helper methods to help you declare
the expectation in an array literal and then verify.

[pdr., 2014-11-14 00:16:51]

PTAL

On 2014/11/13 at 18:42:47, chrishtr wrote:
>
https://codereview.chromium.org/719353004/diff/40001/Source/core/paint/ViewDi...
> File Source/core/paint/ViewDisplayListTest.cpp (right):
> 
>
https://codereview.chromium.org/719353004/diff/40001/Source/core/paint/ViewDi...
> Source/core/paint/ViewDisplayListTest.cpp:353: EXPECT_EQ((size_t)4,
firstList.size());
> This is tedious and hard to read. Add some helper methods to help you declare
the expectation in an array literal and then verify.

Fixed. I added a new macro, EXPECT_DISPLAY_LIST, and used it for all
ViewDisplayListTests.

[chrishtr, 2014-11-14 01:11:26]

lgtm

https://codereview.chromium.org/719353004/diff/60001/Source/core/paint/ViewDi...
File Source/core/paint/ViewDisplayListTest.cpp (left):

https://codereview.chromium.org/719353004/diff/60001/Source/core/paint/ViewDi...
Source/core/paint/ViewDisplayListTest.cpp:63: // TODO(schenney): Check that the
IDs are what we expect.
Just a regular FIXME..

[pdr., 2014-11-14 01:12:57]

The CQ bit was checked by pdr@chromium.org

[commit-bot: I haz the power, 2014-11-14 01:14:20]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/719353004/60001

[commit-bot: I haz the power, 2014-11-14 02:21:28]

Committed patchset #4 (id:60001) as 185341