1439824 - Drop request animationFrame from reftests that suppose that a Web Animation API call posts a request restyle

Status: RESOLVED INVALID

(Core :: DOM: Animation, enhancement)

Description

[Hiroyuki Ikezoe (:hiro)]

Some reftests in layout/reftests/web-animations (e.g. stacking-context-opacity-changing-effect.html) waits for a requestAnimationFrame before removing 'reftest-wait' class.  But animations in those reftest has the same 'from' and 'to' keyframe values, so no invalidation change happens.  We should drop requestAnimationFrame there.

Comment 1

[Hiroyuki Ikezoe (:hiro)]

No more failures in R8 other than stacking-context-transform-changing-display-property.html.
https://treeherder.mozilla.org/#/jobs?repo=try&revision=f5173d84383751a2d32614b4d718b159b6eb72e8

Comment 2

[Hiroyuki Ikezoe (:hiro)]

Attachment: Bug 1439824 - Drop requestAnimationFrame from reftests that suppose that a Web Animation API call posts a request restyle.

All animations in the reftests have the same 'from' and 'to' values so
invalidation changes only in the first frame.

Review commit: https://reviewboard.mozilla.org/r/221870/diff/#index_header
See other reviews: https://reviewboard.mozilla.org/r/221870/

Comment 3

[Hiroyuki Ikezoe (:hiro)]

This is also one of the blockers.

Comment 4

[Brian Birtles (:birtles)]

What's the issue? That reftest-no-flush waits for a restyle before snapshotting? Do these tests need reftest-no-flush?

I'm not 100% clear what the intent of the tests is and why this change is ok? I'm sure it's probably fine but I'd like to understand a little better.

Comment 5

[Hiroyuki Ikezoe (:hiro)]

OK, these test failures will be fixed by bug 1441713.