Summary: | accessibility/smart-invert-reference.html is a flaky failure, always fails when run after rebooting | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Ryan Haddad <ryanhaddad> | ||||||||||
Component: | Layout and Rendering | Assignee: | Nobody <webkit-unassigned> | ||||||||||
Status: | NEW --- | ||||||||||||
Severity: | Normal | CC: | aaron_chu, aboxhall, apinheiro, ap, bfulgham, cfleizach, dmazzoni, ews-watchlist, jcraig, jdiggs, jlewis3, realdawei, samuel_white, simon.fraser, tsavell, webkit-bug-importer, zalan | ||||||||||
Priority: | P2 | Keywords: | InRadar | ||||||||||
Version: | WebKit Nightly Build | ||||||||||||
Hardware: | Unspecified | ||||||||||||
OS: | Unspecified | ||||||||||||
See Also: |
https://bugs.webkit.org/show_bug.cgi?id=205308 https://bugs.webkit.org/show_bug.cgi?id=234591 |
||||||||||||
Attachments: |
|
Description
Ryan Haddad
2018-09-20 20:58:02 PDT
Created attachment 350320 [details]
EWS results
Created attachment 350321 [details]
expected image
Created attachment 350322 [details]
diff
(In reply to Ryan Haddad from comment #3) > Created attachment 350322 [details] > diff This diff shows that the failure is due to a blur around three play buttons. This test was added with https://trac.webkit.org/changeset/226825 It does not appear to be flaky on trunk bots: https://webkit-test-results.webkit.org/dashboards/flakiness_dashboard.html#showAllRuns=true&tests=accessibility%2Fsmart-invert-reference.html Marked test as flaky in https://trac.webkit.org/r236377 We need to try to reproduce this locally. Created attachment 358283 [details]
Patch
Comment on attachment 358283 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=358283&action=review > LayoutTests/ChangeLog:9 > + Swapped out the blur effect image behind the video play buttons for a solid color image to enhance rendering consistency. Did you figure out what was the cause of flakiness specifically? We don't generally have any randomness in painting. One common case for varying results is state leaking from previous tests, and that's something that would need to be fixed if that's the case here. Comment on attachment 358283 [details]
Patch
Shown found repro steps (see radar), and confirmed that the issue still occurs with this patch applied.
*** Bug 205308 has been marked as a duplicate of this bug. *** |