Currently, we use setVideoFullscreenLayer(PlatformLayer* platformLayer, WTF::Function<void()>&& completionHandler) in media element and media player to enter and exit video fullscreen/picture-in-picture. When the platformLayer is valid, we will enter fullscreen/picture-in-picture. When it is a NULL pointer, we will exit fullscreen/picture-in-picture. We had better add some new interfaces to implement that.