Bandaid patch so that we continue with crosssite navigations instead of closing the tab if the beforeunload /unload handler hangs. This patch does the right user-visible behavior, but I'm not a huge fan of the plumbing necessary to make it work. Totally open to cleanup suggestions.

There's also currently one bug that I haven't been able to pinpoint in the UI test. It only treats the first UI test of the four that I run as a cross-site navigation. No matter which test I run first. I wonder if there is some state I should be setting/clearing before/after each test run?

Also there's a DHECK that we hit that the UI test exposed. I 'm not sure it's a case that a user could actually hit though and it's not new with this code, so I added a TODO.

Can I get help from a mac person on adding the UI test to the xcode project?

BUG=3198
Review URL: http://codereview.chromium.org/8920

git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4855 0039d316-1c4b-4281-b951-d872f2087c98

1 files changed, 9 insertions, 0 deletions

diff --git a/chrome/test/ui/ui_test.cc b/chrome/test/ui/ui_test.cc
index 322f02c..f4371d2 100644
--- a/chrome/test/ui/ui_test.cc
+++ b/chrome/test/ui/ui_test.cc
@@ -342,6 +342,15 @@ TabProxy* UITest::GetActiveTab() {
   return window_proxy->GetTab(active_tab_index);
 }
 
+void UITest::NavigateToURLAsync(const GURL& url) {
+  scoped_ptr<TabProxy> tab_proxy(GetActiveTab());
+  ASSERT_TRUE(tab_proxy.get());
+  if (!tab_proxy.get())
+    return;
+
+  tab_proxy->NavigateToURLAsync(url);
+}
+
 void UITest::NavigateToURL(const GURL& url) {
   scoped_ptr<TabProxy> tab_proxy(GetActiveTab());
   ASSERT_TRUE(tab_proxy.get());