PlzNavigate: have renderer-initiated navigations be same-process

content/browser/frame_host/navigator_impl.cc

Index: content/browser/frame_host/navigator_impl.cc
diff --git a/content/browser/frame_host/navigator_impl.cc b/content/browser/frame_host/navigator_impl.cc
index 72f9f8702c8fa0509b62e2bbdbf6a67661c0bf9e..1bab14d4ba73d7f8d03a53d809fdb2558e368b27 100644
--- a/content/browser/frame_host/navigator_impl.cc
+++ b/content/browser/frame_host/navigator_impl.cc
@@ -853,6 +853,9 @@ void NavigatorImpl::RequestNavigation(
       navigation_request_map_.get(frame_tree_node_id);
   request_to_send->SetWaitingForRendererResponse();
   frame_tree_node->current_frame_host()->DispatchBeforeUnload(true);
+
+  if (delegate_)

[clamy, 2015/03/05 10:44:04]

This does not seem related to the change at hand (unlike adding the proper calls
to SendRendererInitiatedNavigation in the unit tests). Could you have it in a
separate patch?

[carlosk, 2015/03/06 11:55:03]

Yes, I noticed the apparent oddness of this before but I wasn't sure about
removing it. But I just ran a simple test that made this clear:
* Running content_unittests with this fix isolated from the rest results in one
fixed test
* Running content_unittests with all other changes but this fix results in all
other expected test to succeed

And this concludes its effects are completely isolated. I'll create a new CL for
this one.

+    delegate_->DidStartNavigationToPendingEntry(entry.GetURL(), reload_type);
 }
 
 void NavigatorImpl::BeginNavigation(FrameTreeNode* frame_tree_node) {