Make sure z-index is adjusted after applying pending animations on elements

Make sure z-index is adjusted after applying pending animations on elements

Before this CL a transform animation that has interpolated transform:none on
the first frame won't have its z-index adjusted to 0 because new animations on
an element are stashed until style finished resolving. If new transform
animations are added then we need to adjust z-index accordingly.

BUG=616674

[trchen, 2016-06-03 23:03:20]

Description was changed from

==========
Make sure z-index is adjusted after applying pending animations on elements

Before this CL a transform animation that has interpolated transform:none on
the first frame won't have its z-index adjusted to 0 because new animations on
an element are stashed until style finished resolving. If new transform
animations are added then we need to adjust z-index accordingly.

BUG=616674
==========

to

==========
Make sure z-index is adjusted after applying pending animations on elements

Before this CL a transform animation that has interpolated transform:none on
the first frame won't have its z-index adjusted to 0 because new animations on
an element are stashed until style finished resolving. If new transform
animations are added then we need to adjust z-index accordingly.

BUG=616674
==========

[trchen, 2016-06-03 23:03:20]

trchen@chromium.org changed reviewers:
+ alancutter@chromium.org, esprehn@chromium.org

[trchen, 2016-06-03 23:05:39]

I tried a few other approaches (forcing interpolated transform:none to
transform:identity(), splitting computed and used z-index) but turns out they
are quite difficult to get right. I resorted to the ugly monkey patch. :(

Do you have other suggestions that I can try? Also do you prefer me to change
the comment from a note to a TODO?

[trchen, 2016-06-03 23:05:46]

The CQ bit was checked by trchen@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2016-06-03 23:06:31]

Dry run: CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/2035793007/1

[commit-bot: I haz the power, 2016-06-04 01:28:01]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2016-06-04 01:28:02]

Dry run: This issue passed the CQ dry run.

[alancutter (OOO until 2018), 2016-06-08 09:59:11]

I'm not too pleased about spreading StyleAdjuster code around.
Would an approach like this work? Make isStackingContext a flag on ComputedStyle
and cease adjusting z-index: https://codereview.chromium.org/2047283002
I don't know enough about layout/paint to make the tests pass or understand
whether it's feasible.

[trchen, 2016-06-08 20:41:13]

On 2016/06/08 09:59:11, alancutter wrote:
> I'm not too pleased about spreading StyleAdjuster code around.
> Would an approach like this work? Make isStackingContext a flag on
ComputedStyle
> and cease adjusting z-index: https://codereview.chromium.org/2047283002
> I don't know enough about layout/paint to make the tests pass or understand
> whether it's feasible.

I think your fix will work. Just worried that there may be code in layout/paint
that invokes hasAutoZIndex when it really means isStackingContext.