[site isolation] cross-site transfers should track the RenderFrameHost, not the View

site isolation: cross-site transfers should track the RenderFrameHost, not the View

This fixes some issues that came up while debugging bug 400850, which inadvertently caused OOPIF-like view reuse in RenderFrameHostManagerTest.AllowTargetedNavigationsAfterSwap.

1. Make navigation_suspended state be per-RenderFrameHost. What was happening, was that we'd suspend navigations on the non-main frame of the view, and upon resume, we'd switch frames.

2. Make the map of pending cross site requests be a map of RenderFrameHost IDs instead of RenderViewHost IDs, instead of just assuming that it's the main frame of the view that's navigating.

3. Add a pending_main_rfh() to the test apparatus, for use by unittests that query the navigation_suspended state.

BUG=304341, 400850
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=288867

[ncarter (slow), 2014-08-09 07:20:59]

Charlie, please have a look

[Charlie Reis, 2014-08-11 17:35:16]

Very nice.  LGTM!

[ncarter (slow), 2014-08-11 17:42:36]

The CQ bit was checked by nick@chromium.org

[commit-bot: I haz the power, 2014-08-11 17:45:02]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/nick@chromium.org/457003002/60001

[commit-bot: I haz the power, 2014-08-12 01:45:00]

Change committed as 288867

[pkotwicz, 2014-08-12 19:22:43]

A revert of this CL has been created in
https://codereview.chromium.org/462083003/ by pkotwicz@chromium.org.

The reason for reverting is: Speculatively reverting CL in case it is causing
WebNavigationApiTest.OpenTab to fail
http://test-results.appspot.com/dashboards/flakiness_dashboard.html#group=%40....