LayoutObject::paintInvalidationParent() to fix multicol spanner invalidation

LayoutObject::paintInvalidationParent() to fix multicol spanner invalidation

This was LayoutObject::parentCrossingFrameBoundaries(). Renamed and
added special rule for column spanner, and used for both finding
painting layer and marking ancestor chain for paint invalidation.

Previously we marked ancestor chain for paint invalidation through
the normal tree order, causing the placeholder not marked for
paint invalidation and missing walking to the spanner. If dcheck
was enabled, it would crash because the paint invalidation flags
of the spanner is not cleared.

BUG=619263
TEST=fast/repaint/multicol-spanner-background-color.html

Committed: https://crrev.com/9c97b32bc9b60b0ec26da4954895ad7eb38d5e79
Cr-Commit-Position: refs/heads/master@{#399614}

[Xianzhu, 2016-06-11 02:06:21]

wangxianzhu@chromium.org changed reviewers:
+ chrishtr@chromium.org, eae@chromium.org

[Xianzhu, 2016-06-13 19:45:00]

ping...

[chrishtr, 2016-06-13 19:49:39]

https://codereview.chromium.org/2063543002/diff/20001/third_party/WebKit/Sour...
File third_party/WebKit/Source/core/layout/LayoutObject.cpp (right):

https://codereview.chromium.org/2063543002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/core/layout/LayoutObject.cpp:2622: if (isLayoutView())
Is it important for this method to not be virtual, and hence have
type-introspection in it?

[Xianzhu, 2016-06-13 19:58:02]

https://codereview.chromium.org/2063543002/diff/20001/third_party/WebKit/Sour...
File third_party/WebKit/Source/core/layout/LayoutObject.cpp (right):

https://codereview.chromium.org/2063543002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/core/layout/LayoutObject.cpp:2622: if (isLayoutView())
On 2016/06/13 19:49:39, chrishtr wrote:
> Is it important for this method to not be virtual, and hence have
> type-introspection in it?

Not important, but just want to make the fact that the function is crossing
frame boundary more explicit here. Also the added logic about isColumnSpanAll()
does belong here because we don't have a class for columnSpanAll.

[chrishtr, 2016-06-13 20:09:33]

lgtm

https://codereview.chromium.org/2063543002/diff/20001/third_party/WebKit/Sour...
File third_party/WebKit/Source/core/layout/LayoutObject.cpp (right):

https://codereview.chromium.org/2063543002/diff/20001/third_party/WebKit/Sour...
third_party/WebKit/Source/core/layout/LayoutObject.cpp:2622: if (isLayoutView())
On 2016/06/13 at 19:58:01, Xianzhu wrote:
> On 2016/06/13 19:49:39, chrishtr wrote:
> > Is it important for this method to not be virtual, and hence have
> > type-introspection in it?
> 
> Not important, but just want to make the fact that the function is crossing
frame boundary more explicit here. Also the added logic about isColumnSpanAll()
does belong here because we don't have a class for columnSpanAll.

It feels a little cleaner to go with virtual like various other methods, but
centralizing logic
is also nice. Up to you.

[Xianzhu, 2016-06-13 20:44:04]

The CQ bit was checked by wangxianzhu@chromium.org

[commit-bot: I haz the power, 2016-06-13 20:45:09]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2063543002/20001

[commit-bot: I haz the power, 2016-06-14 00:22:13]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2016-06-14 00:22:17]

CQ bit was unchecked

[commit-bot: I haz the power, 2016-06-14 00:25:26]

Description was changed from

==========
LayoutObject::paintInvalidationParent() to fix multicol spanner invalidation

This was LayoutObject::parentCrossingFrameBoundaries(). Renamed and
added special rule for column spanner, and used for both finding
painting layer and marking ancestor chain for paint invalidation.

Previously we marked ancestor chain for paint invalidation through
the normal tree order, causing the placeholder not marked for
paint invalidation and missing walking to the spanner. If dcheck
was enabled, it would crash because the paint invalidation flags
of the spanner is not cleared.

BUG=619263
TEST=fast/repaint/multicol-spanner-background-color.html
==========

to

==========
LayoutObject::paintInvalidationParent() to fix multicol spanner invalidation

This was LayoutObject::parentCrossingFrameBoundaries(). Renamed and
added special rule for column spanner, and used for both finding
painting layer and marking ancestor chain for paint invalidation.

Previously we marked ancestor chain for paint invalidation through
the normal tree order, causing the placeholder not marked for
paint invalidation and missing walking to the spanner. If dcheck
was enabled, it would crash because the paint invalidation flags
of the spanner is not cleared.

BUG=619263
TEST=fast/repaint/multicol-spanner-background-color.html

Committed: https://crrev.com/9c97b32bc9b60b0ec26da4954895ad7eb38d5e79
Cr-Commit-Position: refs/heads/master@{#399614}
==========

[commit-bot: I haz the power, 2016-06-14 00:25:27]

Patchset 2 (id:??) landed as
https://crrev.com/9c97b32bc9b60b0ec26da4954895ad7eb38d5e79
Cr-Commit-Position: refs/heads/master@{#399614}