Bug 89301 – [EFL] media/video-seek-past-end-playing.html is flaky

• Status: RESOLVED
• Resolution: FIXED
• Component: WebKit EFL

[Chris Dumez]

media/video-seek-past-end-playing.html is flaky on EFL bot. The diff looks like:

--- /home/buildslave-1/webkit-buildslave/efl-linux-64-debug/build/layout-test-results/media/video-seek-past-end-playing-expected.txt
+++ /home/buildslave-1/webkit-buildslave/efl-linux-64-debug/build/layout-test-results/media/video-seek-past-end-playing-actual.txt
@@ -10,7 +10,7 @@
 RUN(video.currentTime = 500)
 
 EXPECTED (video.paused == 'false') OK
-EXPECTED (mediaElement.currentTime > '0') OK
+EXPECTED (mediaElement.currentTime > '0'), OBSERVED '0' FAIL
 
 END OF TEST

[Chris Dumez]

Skipping test in Bug 89302.

[Chris Dumez]

Created attachment 148095 [details]
Patch

I have made sure that the patch does not cause any regression in both EFL and GTK ports.

[Philippe Normand]

Comment on attachment 148095 [details]
Patch

Please check this patch doesn't break frame accurate seeking. One demo: http://www.massive-interactive.nl/html5_video/smpte_test_universal.html

[WebKit Review Bot]

Comment on attachment 148095 [details]
Patch

Attachment 148095 [details] did not pass chromium-ews (chromium-xvfb):
Output: http://queues.webkit.org/results/12982078

New failing tests:
fast/text/international/thai-line-breaks.html
fast/loader/unload-form-post-about-blank.html

[WebKit Review Bot]

Created attachment 148100 [details]
Archive of layout-test-results from ec2-cr-linux-03

The attached test failures were seen while running run-webkit-tests on the chromium-ews.
Bot: ec2-cr-linux-03  Port: <class 'webkitpy.common.config.ports.ChromiumXVFBPort'>  Platform: Linux-2.6.35-28-virtual-x86_64-with-Ubuntu-10.10-maverick

[Philippe Normand]

Comment on attachment 148095 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=148095&action=review

> Source/WebCore/platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:-1359
> -    // EOS was reached but the position is not always 0 in case of
> -    // reverse playback (or the same as duration in case of forward

That issue might still be valid (position > 0 in reverse playback). IIRC it happens with ogg/theora videos, I think I reported a gst bug for oggdemux about this.

[Chris Dumez]

Created attachment 148106 [details]
Patch

Update the patch based on Philippe's comment. I added back the workaround for the reverse playing case (in case it is still needed and not tested).

My patch does not appear to cause any problem with http://www.massive-interactive.nl/html5_video/smpte_test_universal.html

[Philippe Normand]

Comment on attachment 148106 [details]
Patch

Looks good, thanks!

[WebKit Review Bot]

Comment on attachment 148106 [details]
Patch

Clearing flags on attachment: 148106

Committed r120654: <http://trac.webkit.org/changeset/120654>

[WebKit Review Bot]

All reviewed patches have been landed.  Closing bug.