VideoControlsManager.VideoControlsManagerMultipleVideosWithAudioAndAutoplay has been a flaky timeout on macOS Sierra. The test first failed with this build: https://build.webkit.org/builders/Apple%20Sierra%20Release%20WK1%20%28Tests%29/builds/3552 https://build.webkit.org/builders/Apple%20Sierra%20Release%20WK1%20%28Tests%29/builds/3552/steps/run-api-tests/logs/stdio The test became flaky after the revision: https://trac.webkit.org/changeset/220053/webkit However I don't think this was the cause.
<rdar://problem/33762612>
Spoke with Jer about the Timeout and determined the regression point is most likely https://trac.webkit.org/changeset/220052/webkit and likely deals with the change in html/HTMLMediaElement.cpp
Skipped this flaky test in https://trac.webkit.org/changeset/220711/webkit
Created attachment 332028 [details] Patch
Comment on attachment 332028 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=332028&action=review > Tools/TestWebKitAPI/Tests/WebKitCocoa/VideoControlsManager.mm:157 > +TEST(VideoControlsManager, VideoControlsManagerMultipleVideosWithAudioAndAutoplay) Looks like this test still fails on Mac.
Comment on attachment 332028 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=332028&action=review > Tools/ChangeLog:9 > + Add the same "twice through the event loop" fix for this tast as was done in r221589. s/tast/task/ > Tools/TestWebKitAPI/Tests/WebKitCocoa/large-videos-with-audio-autoplay.html:7 > + // to happen in a certain number of timeouts. Is there a way to make Internals API know whether controls are visible? If so, we could inject Internals in that test instead of using setTimeout. > Tools/TestWebKitAPI/Tests/WebKitCocoa/large-videos-with-audio-autoplay.html:10 > + webkit.messageHandlers.testHandler.postMessage("autoplaying"); Should it be "autoplayed" instead of "autoplaying"?
Comment on attachment 332028 [details] Patch This has been requesting review for more than one year. If this is still needed, please rebase and re-request review.