[CSS Blending] Fix blending when mix-blend-mode is set from javascript

[CSS Blending] Fix blending when mix-blend-mode is set from javascript

If mix-blend-mode is set from javascript to an element that is also a
stacking context, blending is not limited to the parent stacking
context.

BUG=229615
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=162763

[rosca, 2013-11-27 08:48:05]

Stephen, can you please review this fix?
Thanks.

[Stephen Chennney, 2013-11-27 13:06:24]

https://codereview.chromium.org/90983002/diff/1/Source/core/rendering/RenderL...
File Source/core/rendering/RenderLayer.cpp (right):

https://codereview.chromium.org/90983002/diff/1/Source/core/rendering/RenderL...
Source/core/rendering/RenderLayer.cpp:536: if (parent() && (!hadBlendMode ||
!hasBlendMode()))
My only concern is that, without a parent, should you still be executing the
former code (that is, calling dirtyAncestor... on this)?

[rosca, 2013-11-27 13:34:08]

https://codereview.chromium.org/90983002/diff/1/Source/core/rendering/RenderL...
File Source/core/rendering/RenderLayer.cpp (right):

https://codereview.chromium.org/90983002/diff/1/Source/core/rendering/RenderL...
Source/core/rendering/RenderLayer.cpp:536: if (parent() && (!hadBlendMode ||
!hasBlendMode()))
On 2013/11/27 13:06:26, Stephen Chenney wrote:
> My only concern is that, without a parent, should you still be executing the
> former code (that is, calling dirtyAncestor... on this)?
If we change the blendMode for an element we should mark as dirty
m_childLayerHasBlendMode for all its ancestors up to the first stacking context.
If it has no ancestor, we'll call setAncestorChainBlendedDescendant [1] at the
time  this element gets added as a child, so that the new ancestors know that
they have a child with blendMode.
The former code was dirtying m_childLayerHasBlendMode for the blending element
itself (B), so it would have recalculated the blending state of B's descendants,
which didn't change actually.

[1]
https://code.google.com/p/chromium/codesearch#chromium/src/third_party/WebKit...

[Stephen Chennney, 2013-11-27 14:02:03]

I get it now. LGTM.

[commit-bot: I haz the power, 2013-11-27 14:12:31]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/rosca@adobe.com/90983002/1

[commit-bot: I haz the power, 2013-11-27 15:08:26]

Change committed as 162763