REGRESSION(r83561): doesn't pause in-window playback during fullscreen playback

• Component: Media
• Status: RESOLVED FIXED

[Philippe Normand]

The videoValve is not linked if fpsdisplaysink is used.

[Philippe Normand]

Created attachment 89183 [details]
proposed patch

[Martin Robinson]

Comment on attachment 89183 [details]
proposed patch

View in context: https://bugs.webkit.org/attachment.cgi?id=89183&action=review

Is it possible to include a test with this change?

> Source/WebCore/platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:1714
>      if (!m_fpsSink) {
>          gst_bin_add(GST_BIN(m_videoSinkBin), m_webkitVideoSink);
> +        actualVideoSink = m_webkitVideoSink;
> +    }
> +

I think it makes sense to turn if (!m_fpsSink) into an else statement here.

[Philippe Normand]

(In reply to comment #2)
> (From update of attachment 89183 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=89183&action=review
> 
> Is it possible to include a test with this change?
> 
> > 

I don't think this can be tested. The fact that we have 2 video-sinks is internal to the private media-player :/

Source/WebCore/platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:1714
> >      if (!m_fpsSink) {
> >          gst_bin_add(GST_BIN(m_videoSinkBin), m_webkitVideoSink);
> > +        actualVideoSink = m_webkitVideoSink;
> > +    }
> > +
> 
> I think it makes sense to turn if (!m_fpsSink) into an else statement here.

OK. I can attach a new patch or do that tweak before landing if you r+ this.

[Philippe Normand]

(In reply to comment #2)
> (From update of attachment 89183 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=89183&action=review
> 
> Is it possible to include a test with this change?
> 
> > Source/WebCore/platform/graphics/gstreamer/MediaPlayerPrivateGStreamer.cpp:1714
> >      if (!m_fpsSink) {
> >          gst_bin_add(GST_BIN(m_videoSinkBin), m_webkitVideoSink);
> > +        actualVideoSink = m_webkitVideoSink;
> > +    }
> > +
> 
> I think it makes sense to turn if (!m_fpsSink) into an else statement here.

Hum sorry but I don't think so, finally :)
m_fpsSink can be set to 0 in 2 cases just above this code, in the if (m_fpsSink) test.

Turning this if (!m_fpsSink) to a else would require to duplicate the code, I think.

[Martin Robinson]

(In reply to comment #3)
> (In reply to comment #2)
> > (From update of attachment 89183 [details] [details])
> > View in context: https://bugs.webkit.org/attachment.cgi?id=89183&action=review
> > Is it possible to include a test with this change?
> I don't think this can be tested. The fact that we have 2 video-sinks is internal to the private media-player :/

Alright. Please include a line in the ChangeLog explaining the lack of test, per typical WebCore ChangeLog style. :)

[Philippe Normand]

Committed r83741: <http://trac.webkit.org/changeset/83741>