Bug 108925
| Summary: | REGRESSION (r141695): http/tests/media/video-play-stall.html is timing out | ||
|---|---|---|---|
| Product: | WebKit | Reporter: | Zan Dobersek <zan> |
| Component: | WebKitGTK | Assignee: | Nobody <webkit-unassigned> |
| Status: | REOPENED | ||
| Severity: | Normal | CC: | chavarria1991, clopez, mrobinson, pnormand, rego |
| Priority: | P2 | Keywords: | Gtk, LayoutTestFailure |
| Version: | 528+ (Nightly build) | ||
| Hardware: | Unspecified | ||
| OS: | Unspecified | ||
Zan Dobersek
The http/tests/media/video-play-stall.html layout test started timing out intermittently after r141695. When the test doesn't time out, it takes 30 seconds to fail (which is the expected outcome of the test).
http://trac.webkit.org/changeset/141695
http://test-results.appspot.com/dashboards/flakiness_dashboard.html#group=%40ToT%20-%20webkit.org&tests=http%2Ftests%2Fmedia%2Fvideo-play-stall.html
| Attachments | ||
|---|---|---|
| Add attachment proposed patch, testcase, etc. |
Philippe Normand
I guess somehow the url is wrongly set, that could explain the timeout. I'm a bit busy now, maybe later today I'll find time to debug this.
Manuel Rego Casasnovas
I've been taking a look to the issue and I can confirm that with r141695 and r141697 (even without it) the test timeouts instead of simply fail. And without r141695 the test simply fails but id doesn't timeout.
The URL seems to be properly set to:
http://127.0.0.1:8000/resources/load-and-stall.cgi?name=../../../media/content/test.mp4&mimeType=video/mp4&stallAt=100000&stallFor=6
At least at the end of WebKitWebSourceGStreamer::webKitWebSrcSetUri the URL is right.
At the end of MediaPlayerPrivateGStreamer::setPlaybinURL the URL has "webkit+" prefix, but it seems the expected value.
Sorry, but I can't help too much here :-(
Philippe Normand
This patch breaks HTTP on-disk buffering. I think part of the issue is in GStreamer's uridecodebin which checks the decoder uri instead of querying the uri to the source element.
Philippe Normand
Closing as bug 108088 was reopened.
Marcos ChavarrÃa Teijeiro (irc: chavaone)
The test is failing after r179220 (http://trac.webkit.org/changeset/179220)
the diff is the following:
--- /home/ch01/wk-tools/layout-test-results/http/tests/media/video-play-stall-expected.txt
+++ /home/ch01/wk-tools/layout-test-results/http/tests/media/video-play-stall-actual.txt
@@ -4,11 +4,9 @@
EVENT(durationchange)
EVENT(loadedmetadata)
EVENT(loadeddata)
-EVENT(canplay)
-RUN(video.play())
EVENT(stalled)
-EXPECTED (video.currentTime != '0') OK
+EXPECTED (video.currentTime != '0'), OBSERVED '0' FAIL
EXPECTED (video.playbackRate === '1') OK
-EXPECTED (video.paused === 'false') OK
+EXPECTED (video.paused === 'false'), OBSERVED 'true' FAIL
END OF TEST
TestExpectations file updated:
webkit.org/b/108925 http/tests/media/video-play-stall.html [ Failure Timeout ]
Carlos Alberto Lopez Perez
Re-opening issue.
This layout test started to time-out again on GTK on the interval [r274292-r274297]
I guess it may be r274293 but we need to test it to be sure
Carlos Alberto Lopez Perez
*** Bug 224070 has been marked as a duplicate of this bug. ***
Philippe Normand
https://results.webkit.org/?suite=layout-tests&test=http%2Ftests%2Fmedia%2Fvideo-play-stall.html points to: https://trac.webkit.org/changeset/274292/webkit
The test was expected text failing and was unflagged, times out since then.