Enable new multicol for testing and experimental web platform features.

Enable new multicol for testing and experimental web platform features.

There's already a RuntimeEnabledFeatures flag for region based
(aka. "new") multicol. So far it's only been used to toggle support
for column-fill (only supported in the new implementation), but will
from now on be used to determine whether to use the region based
implementation or not.

Remove internals.settings.setRegionBasedColumnsEnabled().

We need to keep a method for region based multicol (that does nothing)
in WebSettings for now, since it's still being called from the
Chromium side. Will be cleaned up later, when the
--enable-region-based-columns command line flag (which from now on has
no effect) is gone. See crbug.com/350853

Remove virtual/regionbasedcolumns tests, since region based multicol
is what's tested by default from now on. At the same time, add
virtual/stable/ tests, so that we still get test coverage for the old
implementation, since that's still what's shipping. This move causes
many changes in TestExpectations. All multicol test expectations that
used to be foo/* will now be virtual/stable/foo/* , and everything
that used to be virtual/regionbased/foo/* is now foo/* .

A few test expectations also need to be updated:

fast/dom/Element/getBoundingClientRect.html: Regenerated expectation
file. New multicol has improved column balancing, resulting in 6px
shorter columns.

fast/borders/border-antialiasing.html: Schedule for
rebaseline. Clipping improvements in new multicol.

fast/repaint/multicol-as-paint-container.html: Schedule for
rebaseline. Different individual rectangles, but same resulting
rectangle.

fast/repaint/multicol-repaint.html: Schedule for rebaseline. Clipping
improvements in new multicol.

BUG=334335, 467468
R=dsinclair@chromium.org,jchaffraix@chromium.org,kbr@chromium.org

Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=192683

[mstensho (USE GERRIT), 2015-03-24 11:02:30]

We're getting ready for new multicol! First we need to land
https://codereview.chromium.org/1024023002/ , though.

There's also one new failure that I need to figure out:
fast/multicol/hit-test-above-or-below.html

It used to pass in both implementations, but now it fails in the new one. I'm
suspecting that platform behavioral differences in text selection are involved
here, but then it's mighty strange that it used to pass.

So this CL isn't quite ready for landing, due to the above reasons. Also,
there's an imminent birth of child for me, which means that I'll disappear for
two weeks any time soon. And since I'd like to be around when and after this
thing has landed, it will have to wait.

But now is a good time to ask for some feedback. :)

This CL is big, due to a huge number of test expectations moving from default to
virtual/stable/, but I decided against splitting it into multiple CLs, as I
think that would be more messy. I think this is something that should be done as
atomically as possible.

[Ken Russell (switch to Gerrit), 2015-03-24 22:04:56]

kbr@chromium.org changed reviewers:
+ esprehn@chromium.org

[Ken Russell (switch to Gerrit), 2015-03-24 22:05:15]

kbr@chromium.org changed reviewers:
- kbr@chromium.org

[esprehn, 2015-03-26 08:30:47]

lgtm, onward and upward! Alf is excited for region based columns.

http://i.imgur.com/vEQp9Xy.gif

[mstensho (USE GERRIT), 2015-03-26 13:36:06]

On 2015/03/26 08:30:47, esprehn wrote:
> lgtm, onward and upward! Alf is excited for region based columns.
> 
> http://i.imgur.com/vEQp9Xy.gif

Woho! It's blocked on the following CLs, though:

https://codereview.chromium.org/1029343003/
https://codereview.chromium.org/1024023002/
https://codereview.chromium.org/1017263004/

[Julien - ping for review, 2015-03-26 15:00:45]

lgtm2 (totally with Alf on this one)

[dsinclair, 2015-03-26 23:17:44]

https://codereview.chromium.org/1028383002/diff/30001/LayoutTests/fast/dom/El...
File LayoutTests/fast/dom/Element/getBoundingClientRect-expected.txt (right):

https://codereview.chromium.org/1028383002/diff/30001/LayoutTests/fast/dom/El...
LayoutTests/fast/dom/Element/getBoundingClientRect-expected.txt:37: FAIL
rect.height.toFixed(3) should be 374.000. Was 368.000.
Are these expected to change?

[mstensho (USE GERRIT), 2015-03-27 09:08:15]

https://codereview.chromium.org/1028383002/diff/30001/LayoutTests/fast/dom/El...
File LayoutTests/fast/dom/Element/getBoundingClientRect-expected.txt (right):

https://codereview.chromium.org/1028383002/diff/30001/LayoutTests/fast/dom/El...
LayoutTests/fast/dom/Element/getBoundingClientRect-expected.txt:37: FAIL
rect.height.toFixed(3) should be 374.000. Was 368.000.
On 2015/03/26 23:17:44, dsinclair wrote:
> Are these expected to change?

Yeah (see commit comment), but let's make it a PASS, not a FAIL. :) Duh!

Done.

[mstensho (USE GERRIT), 2015-03-27 09:08:48]

The CQ bit was checked by mstensho@opera.com

[mstensho (USE GERRIT), 2015-03-27 09:08:50]

The patchset sent to the CQ was uploaded after l-g-t-m from esprehn@chromium.org
Link to the patchset: https://codereview.chromium.org/1028383002/#ps60001
(title: "Mark DRT test svg/text/columns-do-not-apply.html as failing. Need to
land this CL before fixing it.")

[commit-bot: I haz the power, 2015-03-27 09:09:06]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1028383002/60001

[mstensho (USE GERRIT), 2015-03-27 12:20:11]

The CQ bit was unchecked by mstensho@opera.com

[mstensho (USE GERRIT), 2015-03-27 12:24:00]

The CQ bit was checked by mstensho@opera.com

[mstensho (USE GERRIT), 2015-03-27 12:24:02]

The patchset sent to the CQ was uploaded after l-g-t-m from esprehn@chromium.org
Link to the patchset: https://codereview.chromium.org/1028383002/#ps70001
(title: "Need to mark virtual/slimmingpaint/svg/text/columns-do-not-apply.html
as failing too.")

[commit-bot: I haz the power, 2015-03-27 12:24:18]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1028383002/70001

[commit-bot: I haz the power, 2015-03-27 15:18:34]

Committed patchset #8 (id:70001) as
https://src.chromium.org/viewvc/blink?view=rev&revision=192683

[mstensho (USE GERRIT), 2015-04-20 20:21:27]

Note: This got reverted some weeks ago:
https://codereview.chromium.org/1050243002