RIM PR: 116205 We should pass stream client Id to platform player when loading a media url, which is needed by BlackBerry::Platform::MediaSSLHandlerStream for handling the certificate failure if media url is a https url and has a certificate failure reported by mm-renderer.
Created attachment 156840 [details] Patch
Comment on attachment 156840 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=156840&action=review td;dr > Source/WebCore/platform/graphics/blackberry/MediaPlayerPrivateBlackBerry.cpp:160 > + HTMLMediaElement* element = static_cast<HTMLMediaElement*>(m_webCorePlayer->mediaPlayerClient()); that is a layer violation > Source/WebCore/platform/graphics/blackberry/MediaPlayerPrivateBlackBerry.cpp:166 > + int playerId = static_cast<FrameLoaderClientBlackBerry*>(topdoc->frame()->loader()->client())->playerId(); that is a BIG layer violation
MediaPlayerPrivateBlackBerry already does both of those in other functions. Can we clean up all the layering violations in a followup patch?
Comment on attachment 156840 [details] Patch Per above.
(In reply to comment #3) > MediaPlayerPrivateBlackBerry already does both of those in other functions. Can we clean up all the layering violations in a followup patch? I r+'d earlier several bugs with layering violations with the agreement that the violations would be cleaned up, eg. https://bugs.webkit.org/show_bug.cgi?id=82930 from mid April. Max filed bug 84291 about one part of the cleanup, but so far it seems that nothing more has been done.
Created attachment 164831 [details] Patch
Comment on attachment 164831 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=164831&action=review > Source/WebCore/platform/graphics/blackberry/MediaPlayerPrivateBlackBerry.cpp:169 > - m_platformPlayer->load(modifiedUrl.utf8().data(), m_webCorePlayer->userAgent().utf8().data(), cookiePairs.utf8().data()); > + m_platformPlayer->load(playerID, modifiedUrl.utf8().data(), m_webCorePlayer->userAgent().utf8().data(), cookiePairs.utf8().data()); > else > - m_platformPlayer->load(modifiedUrl.utf8().data(), m_webCorePlayer->userAgent().utf8().data(), 0); > + m_platformPlayer->load(playerID, modifiedUrl.utf8().data(), m_webCorePlayer->userAgent().utf8().data(), 0); Why did this compile if the number of parameters was incorrect?
Comment on attachment 164831 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=164831&action=review >> Source/WebCore/platform/graphics/blackberry/MediaPlayerPrivateBlackBerry.cpp:169 >> + m_platformPlayer->load(playerID, modifiedUrl.utf8().data(), m_webCorePlayer->userAgent().utf8().data(), 0); > > Why did this compile if the number of parameters was incorrect? Because our porting haven't got the buildbot for now :(, and we work on our private webkit repo firstly, then sync with upstream. This patch intends to pass the playerID into m_platformPlayer in order to create a SSLHandlingStream to perform SSL verification for the media engine.
(In reply to comment #7) > (From update of attachment 164831 [details]) > View in context: https://bugs.webkit.org/attachment.cgi?id=164831&action=review > > > Source/WebCore/platform/graphics/blackberry/MediaPlayerPrivateBlackBerry.cpp:169 > > - m_platformPlayer->load(modifiedUrl.utf8().data(), m_webCorePlayer->userAgent().utf8().data(), cookiePairs.utf8().data()); > > + m_platformPlayer->load(playerID, modifiedUrl.utf8().data(), m_webCorePlayer->userAgent().utf8().data(), cookiePairs.utf8().data()); > > else > > - m_platformPlayer->load(modifiedUrl.utf8().data(), m_webCorePlayer->userAgent().utf8().data(), 0); > > + m_platformPlayer->load(playerID, modifiedUrl.utf8().data(), m_webCorePlayer->userAgent().utf8().data(), 0); > > Why did this compile if the number of parameters was incorrect? The number of parameters used to be correct. Then we updated the BlackBerry::Platform::PlatformPlayer class in the blackberry "platform" library (an OS service), so now webkit won't compile with the most recent version of the platform lib. This patch updates webkit to compile with the latest version of "platform".
Comment on attachment 164831 [details] Patch rs=me given Joe's informal review.
Comment on attachment 164831 [details] Patch Clearing flags on attachment: 164831 Committed r129244: <http://trac.webkit.org/changeset/129244>
All reviewed patches have been landed. Closing bug.