Bug 223104

Summary: [MacOS Wk1] compositing/visibility/iframe-visibility-hidden.html is a flakey image failure
Product: WebKit Reporter: Amir Mark Jr <amir_mark>
Component: CompositingAssignee: Jean-Yves Avenard [:jya] <jean-yves.avenard>
Status: NEW ---    
Severity: Normal CC: ap, calvaris, cdumez, changseok, eric.carlson, esprehn+autocc, ews-watchlist, glenn, gyuyoung.kim, jean-yves.avenard, jer.noble, kondapallykalyan, pdr, peng.liu6, philipj, sergio, simon.fraser, webkit-bot-watchers-bugzilla, webkit-bug-importer
Priority: P2 Keywords: InRadar
Version: WebKit Nightly Build   
Hardware: Unspecified   
OS: Unspecified   
Attachments:
Description Flags
Expected
none
Actual
none
Diff
none
Patch
none
Patch
none
Patch
none
Patch none

Description Amir Mark Jr 2021-03-11 16:54:07 PST 
compositing/visibility/iframe-visibility-hidden.html

Is a flakey failure that is happening on Big Sur Wk1 debug and release, Big Sur Arm64 Wk1 debug and release, and Catalina WK1 debug and release.


History:

https://results.webkit.org/?suite=layout-tests&test=compositing%2Fvisibility%2Fiframe-visibility-hidden.html


Image Diff:

https://build.webkit.org/results/Apple-BigSur-Release-WK1-Tests/r274301%20(948)/compositing/visibility/iframe-visibility-hidden-diff.png
Comment 1 Amir Mark Jr 2021-03-11 16:55:32 PST 
Created attachment 422993 [details]
Expected
Comment 2 Amir Mark Jr 2021-03-11 16:56:09 PST 
Created attachment 422994 [details]
Actual
Comment 3 Amir Mark Jr 2021-03-11 16:56:45 PST 
Created attachment 422995 [details]
Diff
Comment 4 Radar WebKit Bug Importer 2021-03-11 16:57:21 PST 
<rdar://problem/75340005>
Comment 5 Amir Mark Jr 2021-03-11 17:11:47 PST 
Able to bisect this failure to r274264. 

Able to reproduce on r274264 but unable to reproduce on r274263 or r274258, using this command:

run-webkit-tests compositing/visibility/iframe-visibility-hidden.html --iterations 1000 -f -1
Comment 6 Amir Mark Jr 2021-03-11 18:11:51 PST 
Updated test expectations:

https://trac.webkit.org/changeset/274315/webkit
Comment 7 Alexey Proskuryakov 2021-03-12 16:43:12 PST 
This change was all in media code, so a little surprising. But if it's indeed that, it has to be a substantial issue. I think.
Comment 8 Simon Fraser (smfr) 2021-03-15 09:50:59 PDT 
Not really surprising; the test case uses video.
Comment 9 Jer Noble 2021-04-26 16:12:29 PDT 
Test is a image failure, not a timeout; video is accurately firing the 'load' event. If the background isn't painting, that's not a video bug.
Comment 10 Jean-Yves Avenard [:jya] 2021-06-15 01:12:27 PDT 
Created attachment 431412 [details]
Patch

test wip
Comment 11 Jean-Yves Avenard [:jya] 2021-06-15 04:06:40 PDT 
Created attachment 431421 [details]
Patch

test wip #2
Comment 12 Jean-Yves Avenard [:jya] 2021-06-16 00:04:13 PDT 
Created attachment 431523 [details]
Patch

test wip #3
Comment 13 Jean-Yves Avenard [:jya] 2021-06-16 05:19:22 PDT 
Created attachment 431538 [details]
Patch

test wip #4