Test that no crash happens with the CWS error page.

chrome/browser/extensions/process_management_browsertest.cc

 // Copyright (c) 2012 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 <stddef.h>
 
 #include "base/strings/utf_string_conversions.h"
 #include "build/build_config.h"
 #include "chrome/browser/extensions/extension_apitest.h"
 #include "chrome/browser/extensions/extension_service.h"
@@ skipping 372 matching lines @@
   // Verify that we really have the Chrome Web Store app loaded in the Web
   // Contents.
   content::RenderProcessHost* new_process_host =
       web_contents->GetMainFrame()->GetProcess();
   EXPECT_TRUE(extensions::ProcessMap::Get(profile())->Contains(
       extensions::kWebStoreAppId, new_process_host->GetID()));
 
   // Verify that Chrome Web Store is isolated in a separate renderer process.
   EXPECT_NE(old_process_host, new_process_host);
 }
+
+// This problem is similar to crbug.com/622385. It happens when an iframe error

[nasko, 2017/03/09 05:07:22]

nit: https://crbug.com...

[arthursonzogni, 2017/03/09 15:40:52]

Done.

+// page with the Chrome Web Store URL is displayed and the iframe and the
+// embedder reside in the same process. In this case, a renderer is killed with

[nasko, 2017/03/09 05:07:22]

nit: s/embedder/parent document/, s/a renderer/the renderer process/

[arthursonzogni, 2017/03/09 15:40:52]

Done.

+// the RFH_CAN_COMMIT_URL_BLOCKED IPC. This test asserts that no crash happens.

[nasko, 2017/03/09 05:07:22]

RFH_CAN_COMMIT_URL_BLOCKED is not an IPC, but an error code.

[arthursonzogni, 2017/03/09 15:40:52]

Done.

+IN_PROC_BROWSER_TEST_F(ChromeWebStoreProcessTest,
+                       ChromeWebStoreBlockedByFrameSrc) {
+  GURL url = embedded_test_server()->GetURL(
+      "/extensions/webstore_blocked_by_frame_src.html");
+  ui_test_utils::NavigateToURL(browser(), url);

[nasko, 2017/03/09 05:07:22]

How does this test assert that there is no crash? I don't see this test
monitoring explicitly for a renderer process going away.

[arthursonzogni, 2017/03/09 15:40:52]

Currently, there is no crash but a DCHECK I put a week ago that is triggered in
the browser process before the crash:
[194887:194887:0309/145431.362148:FATAL:navigation_request.cc(594)]
Check failed: render_frame_host->CanCommitURL(common_params_.url).

I the next patch, it is checked that the renderer is still alive. FYI, when the
test fails, it fails even if I remove the DCHECK, but this time because of the
renderer has been killed.

+}