Navigations are renderer-initiated if no WebUI bindings present.

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 9e01343db6382930805063b8321ecfb3c1242f72..439c5ec47b481330fc759d1c63175c7b777ad7ea 100644
--- a/content/browser/frame_host/navigator_impl.cc
+++ b/content/browser/frame_host/navigator_impl.cc
@@ -562,24 +562,19 @@ void NavigatorImpl::RequestTransferURL(
   params.should_replace_current_entry = should_replace_current_entry;
   params.user_gesture = user_gesture;
 
-  if (GetRenderManager(render_frame_host)->web_ui()) {
-    // Web UI pages sometimes want to override the page transition type for
-    // link clicks (e.g., so the new tab page can specify AUTO_BOOKMARK for

[Charlie Reis, 2015/07/15 20:38:00]

Looking deeper, this change doesn't look safe.

NewTabUI still calls SetLinkTransitionType with AUTO_BOOKMARK, so the NTP may
still rely on this.  @kmadhusu: Can you confirm?

https://code.google.com/p/chromium/codesearch#chromium/src/chrome/browser/ui/...

I also see it used by the bookmark manager extension in ExtensionWebUI (which
doesn't have bindings):
https://code.google.com/p/chromium/codesearch#chromium/src/chrome/browser/ext...

[kmadhusu, 2015/07/17 04:06:47]

mathp@/fserb@: Please address this comment.

[fserb, 2015/07/17 15:59:11]

Yep. We still rely on this when we analyse history. This should not be removed.

-    // automatically generated suggestions).  We don't override other types
-    // like TYPED because they have different implications (e.g., autocomplete).
-    if (ui::PageTransitionCoreTypeIs(
-        params.transition, ui::PAGE_TRANSITION_LINK))
-      params.transition =
-          GetRenderManager(render_frame_host)->web_ui()->
-              GetLinkTransitionType();
-
-    // Note also that we hide the referrer for Web UI pages. We don't really
+  // Note: extensions don't have bindings, so will not be affected by the

[Charlie Reis, 2015/07/15 20:38:00]

Extensions aren't a concept in content/, so I don't think we need this comment.

+  // following code.
+  int enabled_bindings =
+      render_frame_host->render_view_host()->GetEnabledBindings();
+  if (enabled_bindings & content::BINDINGS_POLICY_WEB_UI) {
+    // Note that we hide the referrer for Web UI pages. We don't really

[Charlie Reis, 2015/07/15 20:37:59]

nit: for Web UI pages with bindings.

     // want web sites to see a referrer of "chrome://blah" (and some
     // chrome: URLs might have search terms or other stuff we don't want to
     // send to the site), so we send no referrer.
     params.referrer = Referrer();

[Charlie Reis, 2015/07/15 20:38:00]

@kalman: Should extension processes be sending a null referrer (as they do today
thanks to this code), or a real referrer?

[not at google - send to devlin, 2015/07/15 20:42:03]

We should maintain the old, existing behavior.

 
-    // Navigations in Web UI pages count as browser-initiated navigations.
+    // Navigations in Web UI pages (with WebUI bindings) count as

[Charlie Reis, 2015/07/15 20:38:00]

nit: Web UI

+    // browser-initiated navigations.
     params.is_renderer_initiated = false;
   }