1840105 - [wpt-sync] Sync PR 40724 - Use the block-in-inline layout result when considering breaks.

Status: RESOLVED FIXED

(Core :: Layout, task, P4)

Description

[Web Platform Test Sync Bot [:wpt-sync] (Matrix: #interop:mozilla.org)]

Sync web-platform-tests PR 40724 into mozilla-central (this bug is closed when the sync is complete).

PR: https://github.com/web-platform-tests/wpt/pull/40724
Details from upstream follow.

Morten Stenshorne <[e-mail]> wrote:

Use the block-in-inline layout result when considering breaks.

The layout result for the line box that's wrapped around the block-
in-inline isn't very useful when it comes to considering break points.
In order to get access to any block break token inside, for instance,
we need the layout result and fragment of the anonymous block, not the
line box. We were making an incorrect decision to break before an
element in MovePastBreakpoint() because of this bug. Since we failed
to realize that the child had actually broken inside (since we didn't
see the break tokens that were actually there), we'd instead check if
the child was taller than the remaining space - which it was. So we
ended up breaking before at a last-resort break point, instead of
using the breakpoint that we had already found inside. This is a rather
bizarre situation, and in this case it's caused by underlying bugs.
But at least we will no longer freeze / crash, with this CL.

We should also be able to remove code added by CL:4181113 from
NGInlineLayoutAlgorithm::PlaceBlockInInline() (setting the
break appeal there manually should no longer be necessary), now that we
use the "right" layout result for fragmentation considerations. But
since this CL is likely to be merged to older versions, let's keep it
as safe and merge-friendly as possible.

NGBoxFragmentBuilder::AddResult() should also call the new function
LayoutResultForPropagation() instead of doing the work on its own, but
for the same reason as above, leave it alone for now.

Bug: 1453643
Change-Id: Iaca8269d46eeaa7dc222298d77873a9c051f71c3

Reviewed-on: https://chromium-review.googlesource.com/4638061
WPT-Export-Revision: e9838eab9a18cf1f56ccca924e8ce00a5a38e2d8

Comment 1

[Web Platform Test Sync Bot [:wpt-sync] (Matrix: #interop:mozilla.org)]

Pushed to try (stability) https://treeherder.mozilla.org/#/jobs?repo=try&revision=50adeab93ba32ad79f2afc036c582205cdfbbb7f

Comment 2

[Web Platform Test Sync Bot [:wpt-sync] (Matrix: #interop:mozilla.org)]

CI Results

Ran 8 Firefox configurations based on mozilla-central, and Firefox, Chrome, and Safari on GitHub CI

Total 2 tests

Status Summary

Firefox

PASS : 2

Chrome

TIMEOUT: 2

Safari

PASS : 2

Links

Gecko CI (Treeherder)
GitHub PR Head
GitHub PR Base

Comment 3

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/autoland/rev/fc62c4096250
[wpt PR 40724] - Use the block-in-inline layout result when considering breaks., a=testonly

Comment 4

[Cristina Horotan [:chorotan]]

https://hg.mozilla.org/mozilla-central/rev/fc62c4096250