Update a comment with a link to a historic source.

content/browser/frame_host/navigation_controller_impl.cc

 // Copyright 2013 The Chromium Authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 #include "content/browser/frame_host/navigation_controller_impl.h"
 
 #include "base/bind.h"
 #include "base/command_line.h"
 #include "base/debug/trace_event.h"
 #include "base/logging.h"
@@ skipping 1197 matching lines @@
   // have to query the entry index again.
   last_committed_entry_index_ =
       GetEntryIndexWithPageID(rfh->GetSiteInstance(), params.page_id);
 }
 
 void NavigationControllerImpl::RendererDidNavigateNewSubframe(
     RenderFrameHost* rfh,
     const FrameHostMsg_DidCommitProvisionalLoad_Params& params) {
   if (ui::PageTransitionCoreTypeIs(params.transition,
                                    ui::PAGE_TRANSITION_AUTO_SUBFRAME)) {
-    // This is not user-initiated. Ignore.
+    // The old comment here was "This is not user-initiated. Ignore." This is a
+    // weird scenario, where we've determined that this was a new subframe
+    // navigation but say "never mind" based on a flag from the renderer.

[Charlie Reis, 2015/01/15 22:12:09]

Maybe I misunderstood this case before, but I thought it was used for the first
navigation in a newly created iframe, which meant it shouldn't create a new
NavigationEntry.  That seems like a pretty standard case to me.

If so, I'm fine with updating the comment to be a bit more explicit about that,
though.

[Avi (use Gerrit), 2015/01/15 22:16:52]

Will investigate.

[Avi (use Gerrit), 2015/01/16 21:06:26]

Isn't that what RendererDidNavigateAutoSubframe is for, though?

I'm rebuilding and looking at this.

+    //
+    // This dates back to http://crrev.com/115919 and the handling of immediate

[Charlie Reis, 2015/01/15 22:12:09]

This revision is a DEPS roll.  Did you mean http://crrev.com/18373 (from
https://codereview.chromium.org/115919)?

That CL is indeed weird and troubling, but thankfully it doesn't look like most
of its time-based ignore-redirect logic is still around.

[Avi (use Gerrit), 2015/01/15 22:16:52]

Yes. I used the auto-shortener and am bummed that it screwed up the URL.
Yeah.

+    // redirects. TODO(avi): Is this still valid?
     DiscardNonCommittedEntriesInternal();
     return;
   }
 
   // Manual subframe navigations just get the current entry cloned so the user
   // can go back or forward to it. The actual subframe information will be
   // stored in the page state for each of those entries. This happens out of
   // band with the actual navigations.
   DCHECK(GetLastCommittedEntry()) << "ClassifyNavigation should guarantee "
                                   << "that a last committed entry exists.";
@@ skipping 548 matching lines @@
     }
   }
 }
 
 void NavigationControllerImpl::SetGetTimestampCallbackForTest(
     const base::Callback<base::Time()>& get_timestamp_callback) {
   get_timestamp_callback_ = get_timestamp_callback;
 }
 
 }  // namespace content