Adds DataChannelInit param to createDataChannel.

content/renderer/media/rtc_peer_connection_handler.h

Index: content/renderer/media/rtc_peer_connection_handler.h
diff --git a/content/renderer/media/rtc_peer_connection_handler.h b/content/renderer/media/rtc_peer_connection_handler.h
index 7fc9288188b1831e90f356c9ec2577cb5817d092..c2bf1b32eddbb7290757f0b65df6bfdad12ee258 100644
--- a/content/renderer/media/rtc_peer_connection_handler.h
+++ b/content/renderer/media/rtc_peer_connection_handler.h
@@ -128,6 +128,9 @@ class CONTENT_EXPORT RTCPeerConnectionHandler
       const WebKit::WebRTCStatsRequest& request) OVERRIDE;
   virtual WebKit::WebRTCDataChannelHandler* createDataChannel(
       const WebKit::WebString& label, bool reliable) OVERRIDE;
+  virtual WebKit::WebRTCDataChannelHandler* createDataChannel(
+      const WebKit::WebString& label,
+      const WebKit::WebRTCDataChannelInit& init) OVERRIDE;

[Ami GONE FROM CHROMIUM, 2013/06/03 18:27:31]

FYI if you remove OVERRIDE you can land CLs a bit more freely (don't have to
wait for as many rolls from chromium into blink and from blink into chromium).
Just remember to add back any OVERRIDEs that apply when you're done.

(chromium has a clang plugin that checks that all virtual methods specify
OVERRIDE, but there's an exception for WebKit-implementing things for exactly
this reason)

[jiayl, 2013/06/03 18:52:58]

Hmm, but I still depend on the definition of DataChannelInit in the same Blink
CL.
On 2013/06/03 18:27:31, Ami Fischman wrote:

[Ami GONE FROM CHROMIUM, 2013/06/03 19:32:19]

True; this would just allow you to remove the old function in the same blink CL
that converts to using the new function.

[jiayl, 2013/06/03 19:54:51]

I see. Makes sense!
On 2013/06/03 19:32:19, Ami Fischman wrote:

   virtual WebKit::WebRTCDTMFSenderHandler* createDTMFSender(
       const WebKit::WebMediaStreamTrack& track) OVERRIDE;
   virtual void stop() OVERRIDE;