REGRESSION(r266634): fast/animation/request-animation-frame-throttling-lowPowerMode.html is flaky

• Component: New Bugs
• Status: RESOLVED FIXED

[Sihui Liu]

...

[Sihui Liu]

Created attachment 408209 [details]
Patch

[Radar WebKit Bug Importer]

<rdar://problem/68513429>

[Darin Adler]

Comment on attachment 408209 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=408209&action=review

> Source/WebKit/ChangeLog:9
> +        Do not schedule rendering update if we don't need to handle activity state change callbacks, as we did not
> +        schedule activity state timer before r266634.

This stops *just* short of saying why we should make this change. I understand that it restore the old behavior and we did not need to change it. But why is the old behavior correct/better? Any why does that help fix "flakiness". Those are left mysterious by this comment.

[Sihui Liu]

Created attachment 408272 [details]
Patch

[Sihui Liu]

(In reply to Darin Adler from comment #3)
> Comment on attachment 408209 [details]
> Patch
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=408209&action=review
> 
> > Source/WebKit/ChangeLog:9
> > +        Do not schedule rendering update if we don't need to handle activity state change callbacks, as we did not
> > +        schedule activity state timer before r266634.
> 
> This stops *just* short of saying why we should make this change. I
> understand that it restore the old behavior and we did not need to change
> it. But why is the old behavior correct/better? Any why does that help fix
> "flakiness". Those are left mysterious by this comment.

Updated ChangeLog after digging into it. Looks like there is actually nothing wrong about the current behavior, so we may also fix this by changing test expectation. Maybe Simon or Said can comment about this.

[EWS]

Committed r266838: <https://trac.webkit.org/changeset/266838>

All reviewed patches have been landed. Closing bug and clearing flags on attachment 408272 [details].