Skip to enter the fullscreen automatically when hole-punching is enabled.

content/browser/media/android/browser_media_player_manager.cc

Index: content/browser/media/android/browser_media_player_manager.cc
diff --git a/content/browser/media/android/browser_media_player_manager.cc b/content/browser/media/android/browser_media_player_manager.cc
index e2d6676d890b5e4d935e2869cb0d8b04e86535f3..3244e9b5e6f22b886f0dc9174a308bb51306ecfb 100644
--- a/content/browser/media/android/browser_media_player_manager.cc
+++ b/content/browser/media/android/browser_media_player_manager.cc
@@ -323,6 +323,12 @@ void BrowserMediaPlayerManager::RequestFullScreen(int player_id) {
   Send(new MediaPlayerMsg_RequestFullscreen(RoutingID(), player_id));
 }
 
+bool
+BrowserMediaPlayerManager::ShouldUseVideoOverlayForEmbeddedEncryptedVideo() {
+  RendererPreferences* prefs = web_contents_->GetMutableRendererPrefs();
+  return prefs->use_video_overlay_for_embedded_encrypted_video;

[xhwang, 2014/07/08 16:23:43]

Is this setting VIDEO_HOLE specific? 

If yes, this setting and this code isn't guarded by VIDEO_HOLE ifdef. Actually
the render side ShouldUseVideoOverlayForEmbeddedEncryptedVideo() code is guarded
by VIDEO_HOLE ifdef.

If no, you need to update the CL description to make this CL not VIDEO_HOLE
specific.

[ycheo (away), 2014/07/09 01:18:07]

Actually, the setting exists regardless of VIDEO_HOLE, but the usage is limited
to VIDEO_HOLE.
I didn't guard this because it makes that the condition in medio_source_player
looks ugly.
But, as you said, the function is only for the VIDEO_HOLE.
I think it would be better to guard it.
 

+}
+
 #if defined(VIDEO_HOLE)
 void BrowserMediaPlayerManager::AttachExternalVideoSurface(int player_id,
                                                            jobject surface) {