Bug 222091

Summary: [GTK] API test /WebKit2Gtk/TestWebKitPolicyClient:/webkit/WebKitPolicyClient/autoplay-policy is flaky timing out on EWS
Product: WebKit Reporter: Lauro Moura <lmoura>
Component: WebKitGTKAssignee: Nobody <webkit-unassigned>
Status: NEW ---    
Severity: Normal CC: bugs-noreply
Priority: P2    
Version: WebKit Nightly Build   
Hardware: Unspecified   
OS: Unspecified   
See Also: https://bugs.webkit.org/show_bug.cgi?id=222090

Description Lauro Moura 2021-02-17 20:21:01 PST
/WebKit2Gtk/TestWebKitPolicyClient:/webkit/WebKitPolicyClient/autoplay-policy

Results from last ~350 builds (~4 days)

* 29762 (igalia3): Flaky
* 29611 (igalia4): Old Failure
* 29578 (igalia4): Flaky
* 29574 (igalia4): Old Failure and Flaky
* 29511 (igalia4): Flaky
* 29449 (igalia4): Flaky

Marking as flaky until investigation on igalia4 is performed (see bug222090 for other test failing there).
Comment 1 Lauro Moura 2021-02-18 19:43:17 PST
Not sure if related to the flakiness on the bots, but I'm getting consistent failures locally in this test too in the first ALLOW_WITHOUT_SOUND check (where no playback is expected).

Digging, there are some mediaPlayerVolumeChanged calls with 0 volume on the player, which zeroes the HTMLMediaElement volume, which in turn is making the audio check in MediaElementSession::playbackPermitted to allow playback regardless of the AllowWithoutSound policy.

Forcing the volume to 1 in the test (autoplay-check.html) makes the test pass.

What could be making the player emit those volumeChanged=0 calls?