Bug 42094

Summary: [Qt] REGRESSION?(62951): media tests fail
Product: WebKit Reporter: Csaba Osztrogonác <ossy>
Component: New BugsAssignee: Nancy Piedra <nancy.piedra>
Status: RESOLVED INVALID    
Severity: Normal CC: abecsi, diegohcg, hausmann, menard, nancy.piedra, pnormand, vestbo, webkit.review.bot
Priority: P2 Keywords: Qt, QtTriaged
Version: 528+ (Nightly build)   
Hardware: All   
OS: All   
Bug Depends on: 86679    
Bug Blocks: 79666    
Attachments:
Description Flags
Patch none

Description Csaba Osztrogonác 2010-07-12 10:43:46 PDT
After r62951 landed, two media tests fail on our bots which 
have Qt 4.6.3 (and Qt 4.7.0) with Qt Mobility API 1.0.1.
They pass on "Qt Linux Release" bot on build.webkit.org,
which has Qt 4.6.2 with Qt Mobility API 1.0.0

Is it a regression? What do you think?

http://webkit.sed.hu/buildbot/results/x86-64%20Linux%20Qt%20Release/r63096%20%286846%29/media/restore-from-page-cache-pretty-diff.html
http://webkit.sed.hu/buildbot/results/x86-64%20Linux%20Qt%20Release/r63096%20%286846%29/media/video-can-play-type-pretty-diff.html

--- /home/webkitbuildbot/slaves/release64bit/buildslave/qt-linux-64-release/build/layout-test-results/media/restore-from-page-cache-expected.txt	2010-07-12 10:32:28.099646483 -0700
+++ /home/webkitbuildbot/slaves/release64bit/buildslave/qt-linux-64-release/build/layout-test-results/media/restore-from-page-cache-actual.txt	2010-07-12 10:32:28.099646483 -0700
@@ -6,4 +6,5 @@
 *** Page going into cache
 *** Page returned from cache
 EXPECTED (loadCount == '1') OK
+EVENT(loadstart)


--- /home/webkitbuildbot/slaves/release64bit/buildslave/qt-linux-64-release/build/layout-test-results/media/video-can-play-type-expected.txt	2010-07-12 10:32:28.164565846 -0700
+++ /home/webkitbuildbot/slaves/release64bit/buildslave/qt-linux-64-release/build/layout-test-results/media/video-can-play-type-actual.txt	2010-07-12 10:32:28.164565846 -0700
@@ -1,10 +1,10 @@
 Test HTMLMediaElement canPlayType() method.
 
 EXPECTED (video.canPlayType() == '') OK
-EXPECTED (video.canPlayType('video/') == '') OK
-EXPECTED (video.canPlayType('video/blahblah') == '') OK
-EXPECTED (video.canPlayType('video/mpeg; codecs="avc1.4D400C"') == 'probably') OK
-EXPECTED (video.canPlayType(' video/mp4 ; video/mp4; codecs="mp4v.20.8, mp4a.40.2"') == 'probably') OK
+EXPECTED (video.canPlayType('video/') == ''), OBSERVED 'maybe' FAIL
+EXPECTED (video.canPlayType('video/blahblah') == ''), OBSERVED 'maybe' FAIL
+EXPECTED (video.canPlayType('video/mpeg; codecs="avc1.4D400C"') == 'probably'), OBSERVED 'maybe' FAIL
+EXPECTED (video.canPlayType(' video/mp4 ; video/mp4; codecs="mp4v.20.8, mp4a.40.2"') == 'probably'), OBSERVED 'maybe' FAIL
 EXPECTED (video.canPlayType('audio/mpeg') == 'maybe') OK
 EXPECTED (video.canPlayType('audio/wav') == 'maybe') OK
 END OF TEST
Comment 1 Nancy Piedra 2011-04-01 05:23:51 PDT
With QtMobility master & Qt 4.7.2 on Linux, restore-from-page-cache passes for me.  But video-can-play-type still fails.  I think that MediaPlayerPrivateQt.cpp does not implement to the HTML5 specification.  I will take a look at that.
Comment 2 Nancy Piedra 2011-04-01 18:25:31 PDT
The root cause of the video-can-play-type has been identified as an error in QtMobility in the following bug: https://bugs.webkit.org/show_bug.cgi?id=53275

This bug can probably be closed as a duplicate of 53275.
Comment 3 Nancy Piedra 2011-04-04 05:40:12 PDT
Acutally, this bug should be kept open because still some changes will be needed to the Qt MediaPlayer implementation to pass this test.
Comment 4 Alexis Menard (darktears) 2012-05-16 07:56:43 PDT
Created attachment 142262 [details]
Patch
Comment 5 Alexis Menard (darktears) 2012-05-16 07:58:13 PDT
(In reply to comment #4)
> Created an attachment (id=142262) [details]
> Patch

Let's give a shot and we can skip some of them if needed. Some coverage (395) is better than 0 coverage.
Comment 6 Philippe Normand 2012-05-16 10:37:24 PDT
(In reply to comment #5)
> (In reply to comment #4)
> > Created an attachment (id=142262) [details] [details]
> > Patch
> 
> Let's give a shot and we can skip some of them if needed. Some coverage (395) is better than 0 coverage.

What about the http/tests/media? I suspect some of them pass as well.
Comment 7 Alexis Menard (darktears) 2012-05-16 10:42:07 PDT
(In reply to comment #6)
> (In reply to comment #5)
> > (In reply to comment #4)
> > > Created an attachment (id=142262) [details] [details] [details]
> > > Patch
> > 
> > Let's give a shot and we can skip some of them if needed. Some coverage (395) is better than 0 coverage.
> 
> What about the http/tests/media? I suspect some of them pass as well.

One thing at a time :)
Comment 8 WebKit Review Bot 2012-05-16 13:30:31 PDT
Comment on attachment 142262 [details]
Patch

Clearing flags on attachment: 142262

Committed r117332: <http://trac.webkit.org/changeset/117332>
Comment 9 WebKit Review Bot 2012-05-16 13:30:38 PDT
All reviewed patches have been landed.  Closing bug.
Comment 10 WebKit Review Bot 2012-05-16 15:01:33 PDT
Re-opened since this is blocked by 86679