[RDHI] Refactored blocked_loaders_map_ to key by render frame route id

content/browser/web_contents/web_contents_impl.cc

Index: content/browser/web_contents/web_contents_impl.cc
diff --git a/content/browser/web_contents/web_contents_impl.cc b/content/browser/web_contents/web_contents_impl.cc
index 4ff0a3a18951b3884bacf6315ffc450b21b2da24..9c80f0c99be40d65f24a6423391fb0fd05de4b51 100644
--- a/content/browser/web_contents/web_contents_impl.cc
+++ b/content/browser/web_contents/web_contents_impl.cc
@@ -403,6 +403,7 @@ WebContentsImpl::WebContentsImpl(BrowserContext* browser_context)
 #else
   media_web_contents_observer_.reset(new MediaWebContentsObserver(this));
 #endif
+  loader_io_thread_notifier_.reset(new LoaderIOThreadNotifier(this));
   wake_lock_service_context_.reset(new WakeLockServiceContext(this));
 }
 
@@ -1848,7 +1849,10 @@ void WebContentsImpl::CreateNewWindow(
       // delete the RenderView that had already been created.
       Send(new ViewMsg_Close(route_id));
     }
-    GetRenderViewHost()->GetProcess()->ResumeRequestsForView(route_id);
+    // It's safe to only target the frame because the render process will not
+    // have a chance to create more frames at this point.

[Randy Smith (Not in Mondays), 2016/02/02 20:47:06]

Is this comment still accurate, given that we're recursing by default?  To me it
implies we're only targeting the main frame.

[Charlie Harrison, 2016/02/02 20:51:59]

This code path (and the path that blocks it) still target a single frame
(route), because it's a new window code path and there isn't a frame tree yet.

+    ResourceDispatcherHostImpl::ResumeBlockedRequestsForRouteFromUI(
+        GlobalFrameRoutingId(render_process_id, main_frame_route_id));
     return;
   }