Summary: | [GTK] Several tests crashing on debug bot in (anonymous namespace)::MediaPlayerPrivateGStreamerBase::repaint | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Michael Catanzaro <mcatanzaro> | ||||||||||
Component: | Media | Assignee: | Miguel Gomez <magomez> | ||||||||||
Status: | RESOLVED FIXED | ||||||||||||
Severity: | Normal | CC: | bugs-noreply, commit-queue, magomez, mcatanzaro | ||||||||||
Priority: | P2 | ||||||||||||
Version: | Other | ||||||||||||
Hardware: | PC | ||||||||||||
OS: | Linux | ||||||||||||
Bug Depends on: | 163602 | ||||||||||||
Bug Blocks: | |||||||||||||
Attachments: |
|
Description
Michael Catanzaro
2016-10-16 18:02:31 PDT
(Not updating any expectations for this.) > The GTK+ debug bot is exiting early due to a bunch of crashes in media
> elements code. I wonder if it could be related to r207287 "[GTK] Video
> playback doesn't work properly with accelerated compositing disabled".
Could be. Checking.
Created attachment 291816 [details]
Patch
Comment on attachment 291816 [details] Patch Attachment 291816 [details] did not pass ios-sim-ews (ios-simulator-wk2): Output: http://webkit-queues.webkit.org/results/2304645 New failing tests: media/controls/ipad/close-page-with-picture-in-picture-video-assertion-failure.html Created attachment 291820 [details]
Archive of layout-test-results from ews121 for ios-simulator-wk2
The attached test failures were seen while running run-webkit-tests on the ios-sim-ews.
Bot: ews121 Port: ios-simulator-wk2 Platform: Mac OS X 10.11.6
Comment on attachment 291816 [details] Patch Clearing flags on attachment: 291816 Committed r207409: <http://trac.webkit.org/changeset/207409> All reviewed patches have been landed. Closing bug. Re-opened since this is blocked by bug 163602 It seems that we can't rely on HTMLMediaElement::mediaPlayerRenderingCanBeAccelerated() to check whether accelerated compositing is enabled or not, as there are situations where this method will return false when acc compositing is enabled. Due to this, with the previous patch some frames were processed through the slow path instead of the acc compositing one, causing the timeouts in some tests. I'm working on a new patch that uses a new method to check whether acc compositing is enabled. This should fix the timeouts. Created attachment 292190 [details]
Patch
Comment on attachment 292190 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=292190&action=review > Source/WebCore/html/HTMLMediaElement.cpp:4584 > + return document().settings()->acceleratedCompositingEnabled(); You should null check settings() Created attachment 292332 [details]
Patch
Comment on attachment 292332 [details] Patch Clearing flags on attachment: 292332 Committed r207666: <http://trac.webkit.org/changeset/207666> All reviewed patches have been landed. Closing bug. It broke imported/w3c/web-platform-tests/html/semantics/embedded-content/media-elements/playing-the-media-resource/pause-move-within-document.html, see bug #163850. (In reply to comment #15) > It broke > imported/w3c/web-platform-tests/html/semantics/embedded-content/media- > elements/playing-the-media-resource/pause-move-within-document.html, see bug > #163850. Shame, it would be good to know why this broke. |