Add browser tests for enabling Google Drive offline mode on first run.

Add browser tests for enabling Google Drive offline mode on first run.

BUG=312466
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=235674

[Tim Song, 2013-11-13 02:42:04]

PTAL

[achuithb, 2013-11-13 04:13:09]

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
File chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc (right):

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc:58:
InProcessBrowserTest::SetUpOnMainThread();
Do you mind pulling out the function bodies to outside the class definition? I
think inlining these makes it harder to read, and I believe the style guide
forbids it except for oneliners.

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc:116:
runner_->Quit();
Should you set controller_ to NULL in this function?

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc:119: virtual
void OnTimedOut() {
OVERRIDE?

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc:123:
DriveFirstRunController* controller() { return controller_; }
Why do you need this?

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc:124: bool
timed_out() { return timed_out_; }
Should these functions be const?

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc:155:
controller()->SetDelaysForTest(0, 0);
Won't the test pass even with a good directory if the timeout is 0?

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
File chrome/browser/chromeos/first_run/drive_first_run_controller.cc (right):

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
chrome/browser/chromeos/first_run/drive_first_run_controller.cc:46: const char
kDriveOfflineEndpointUrl[] = "https://drive.google.com/#offline";
Wonder if this particular change should be split out into a separate CL?

If not, let's call it out in the CL description.

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
chrome/browser/chromeos/first_run/drive_first_run_controller.cc:301: return;
This is a bug fix, right? If you do decide to make a separate CL, this should be
in it.

[Tim Song, 2013-11-13 20:00:25]

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
File chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc (right):

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc:58:
InProcessBrowserTest::SetUpOnMainThread();
On 2013/11/13 04:13:10, achuith.bhandarkar wrote:
> Do you mind pulling out the function bodies to outside the class definition? I
> think inlining these makes it harder to read, and I believe the style guide
> forbids it except for oneliners.

Done. There was a discussion on chromium-dev about inlining functions in test
files:
https://groups.google.com/a/chromium.org/forum/#!msg/chromium-dev/X8D176Uzp6w...

And the style guide has been updated to make an exception for tests if it is
readable.

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc:116:
runner_->Quit();
On 2013/11/13 04:13:10, achuith.bhandarkar wrote:
> Should you set controller_ to NULL in this function?

Done.

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc:119: virtual
void OnTimedOut() {
On 2013/11/13 04:13:10, achuith.bhandarkar wrote:
> OVERRIDE?

Done.

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc:123:
DriveFirstRunController* controller() { return controller_; }
On 2013/11/13 04:13:10, achuith.bhandarkar wrote:
> Why do you need this?

controller_ is private and can't be referenced in the tests.

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc:124: bool
timed_out() { return timed_out_; }
On 2013/11/13 04:13:10, achuith.bhandarkar wrote:
> Should these functions be const? 

Done.

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc:155:
controller()->SetDelaysForTest(0, 0);
On 2013/11/13 04:13:10, achuith.bhandarkar wrote:
> Won't the test pass even with a good directory if the timeout is 0?

That's true, but it would take a few seconds to complete the flow successfully.
I'm not sure it's a good idea to wait that long in a browser test in the
expected case. This test case mainly tests that we get an expected time out and
not a test timeout that would cause an unexpected exit.

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
File chrome/browser/chromeos/first_run/drive_first_run_controller.cc (right):

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
chrome/browser/chromeos/first_run/drive_first_run_controller.cc:46: const char
kDriveOfflineEndpointUrl[] = "https://drive.google.com/#offline";
On 2013/11/13 04:13:10, achuith.bhandarkar wrote:
> Wonder if this particular change should be split out into a separate CL?
> 
> If not, let's call it out in the CL description.

Done. I'll move this to another CL. However, this isn't the final endpoint we'll
be using, as it is pretty heavy-weight.

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
chrome/browser/chromeos/first_run/drive_first_run_controller.cc:301: return;
On 2013/11/13 04:13:10, achuith.bhandarkar wrote:
> This is a bug fix, right? If you do decide to make a separate CL, this should
be
> in it.

Done. Moved to other CL.

[achuithb, 2013-11-13 23:11:54]

lgtm

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
File chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc (right):

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc:155:
controller()->SetDelaysForTest(0, 0);
On 2013/11/13 20:00:25, Tim Song wrote:
> On 2013/11/13 04:13:10, achuith.bhandarkar wrote:
> > Won't the test pass even with a good directory if the timeout is 0?
> 
> That's true, but it would take a few seconds to complete the flow
successfully.
> I'm not sure it's a good idea to wait that long in a browser test in the
> expected case. This test case mainly tests that we get an expected time out
and
> not a test timeout that would cause an unexpected exit.

Could you please add a comment explaining what this test is doing?

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
File chrome/browser/chromeos/first_run/drive_first_run_controller.cc (right):

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
chrome/browser/chromeos/first_run/drive_first_run_controller.cc:46: const char
kDriveOfflineEndpointUrl[] = "https://drive.google.com/#offline";
On 2013/11/13 20:00:25, Tim Song wrote:
> On 2013/11/13 04:13:10, achuith.bhandarkar wrote:
> > Wonder if this particular change should be split out into a separate CL?
> > 
> > If not, let's call it out in the CL description.
> 
> Done. I'll move this to another CL. However, this isn't the final endpoint
we'll
> be using, as it is pretty heavy-weight.

Sure. I just didn't want to mix "real" code with a CL for testing. So if I'm not
mistaken, this CL now has no behavioral changes.

https://codereview.chromium.org/71013003/diff/80001/chrome/browser/chromeos/f...
File chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc (right):

https://codereview.chromium.org/71013003/diff/80001/chrome/browser/chromeos/f...
chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc:68:
DriveFirstRunController* controller() const { return controller_; }
Have you considered having protected methods like
DriveFirstRunTest::EnableOfflineMode, and DriveFirstRunTest::SetDelaysForTest
instead of passing this pointer out?

https://codereview.chromium.org/71013003/diff/80001/chrome/browser/chromeos/f...
chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc:93:
host_resolver()->AddRule("*", "127.0.0.1");
Maybe a comment explaining this? Wouldn't this test work if you used example.com
instead of "*"?

[Tim Song, 2013-11-14 01:18:48]

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
File chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc (right):

https://codereview.chromium.org/71013003/diff/1/chrome/browser/chromeos/first...
chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc:155:
controller()->SetDelaysForTest(0, 0);
On 2013/11/13 23:11:54, achuith.bhandarkar wrote:
> On 2013/11/13 20:00:25, Tim Song wrote:
> > On 2013/11/13 04:13:10, achuith.bhandarkar wrote:
> > > Won't the test pass even with a good directory if the timeout is 0?
> > 
> > That's true, but it would take a few seconds to complete the flow
> successfully.
> > I'm not sure it's a good idea to wait that long in a browser test in the
> > expected case. This test case mainly tests that we get an expected time out
> and
> > not a test timeout that would cause an unexpected exit.
> 
> Could you please add a comment explaining what this test is doing?

Done.

https://codereview.chromium.org/71013003/diff/80001/chrome/browser/chromeos/f...
File chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc (right):

https://codereview.chromium.org/71013003/diff/80001/chrome/browser/chromeos/f...
chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc:68:
DriveFirstRunController* controller() const { return controller_; }
On 2013/11/13 23:11:55, achuith.bhandarkar wrote:
> Have you considered having protected methods like
> DriveFirstRunTest::EnableOfflineMode, and DriveFirstRunTest::SetDelaysForTest
> instead of passing this pointer out?

Done.

https://codereview.chromium.org/71013003/diff/80001/chrome/browser/chromeos/f...
chrome/browser/chromeos/first_run/drive_first_run_browsertest.cc:93:
host_resolver()->AddRule("*", "127.0.0.1");
On 2013/11/13 23:11:55, achuith.bhandarkar wrote:
> Maybe a comment explaining this? Wouldn't this test work if you used
http://example.com
> instead of "*"?

Done. Changed to example.com.

[achuithb, 2013-11-14 01:22:58]

lgtm

[commit-bot: I haz the power, 2013-11-14 18:12:39]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/tengs@chromium.org/71013003/170001

[commit-bot: I haz the power, 2013-11-14 18:12:43]

Failed to apply patch for
chrome/browser/chromeos/first_run/drive_first_run_controller.cc:
While running patch -p1 --forward --force --no-backup-if-mismatch;
  patching file chrome/browser/chromeos/first_run/drive_first_run_controller.cc
  Hunk #2 succeeded at 37 with fuzz 1.
  Hunk #14 FAILED at 295.
  Hunk #15 succeeded at 310 (offset 1 line).
  Hunk #16 succeeded at 319 (offset 1 line).
  Hunk #17 succeeded at 327 (offset 1 line).
  Hunk #18 succeeded at 377 (offset 1 line).
  1 out of 18 hunks FAILED -- saving rejects to file
chrome/browser/chromeos/first_run/drive_first_run_controller.cc.rej

Patch:       chrome/browser/chromeos/first_run/drive_first_run_controller.cc
Index: chrome/browser/chromeos/first_run/drive_first_run_controller.cc
diff --git a/chrome/browser/chromeos/first_run/drive_first_run_controller.cc
b/chrome/browser/chromeos/first_run/drive_first_run_controller.cc
index
dae455b2a62f7b983daeded9c3871eef247a69ae..eef56457d1e42ff7436228b20518c2a01a65c7c2
100644
--- a/chrome/browser/chromeos/first_run/drive_first_run_controller.cc
+++ b/chrome/browser/chromeos/first_run/drive_first_run_controller.cc
@@ -16,6 +16,7 @@
 #include "chrome/browser/chromeos/login/user_manager.h"
 #include "chrome/browser/extensions/extension_service.h"
 #include "chrome/browser/extensions/extension_system.h"
+#include "chrome/browser/extensions/extension_web_contents_observer.h"
 #include "chrome/browser/profiles/profile_manager.h"
 #include "chrome/browser/tab_contents/background_contents.h"
 #include "chrome/common/extensions/extension.h"
@@ -36,10 +37,10 @@ namespace chromeos {
 namespace {
 
 // The initial time to wait in seconds before starting the opt-in.
-const int kInitialDelaySeconds = 180;
+int kInitialDelaySeconds = 180;
 
 // Time to wait for Drive app background page to come up before giving up.
-const int kWebContentsTimeoutSeconds = 15;
+int kWebContentsTimeoutSeconds = 15;
 
 // Google Drive offline opt-in endpoint.
 const char kDriveOfflineEndpointUrl[] = "http://drive.google.com";
@@ -63,6 +64,8 @@ class DriveWebContentsManager : public
content::WebContentsObserver,
   typedef base::Callback<void(bool)> CompletionCallback;
 
   DriveWebContentsManager(Profile* profile,
+                          const std::string& app_id,
+                          const std::string& endpoint_url,
                           const CompletionCallback& completion_callback);
   virtual ~DriveWebContentsManager();
 
@@ -114,6 +117,8 @@ class DriveWebContentsManager : public
content::WebContentsObserver,
                        const content::NotificationDetails& details) OVERRIDE;
 
   Profile* profile_;
+  const std::string app_id_;
+  const std::string endpoint_url_;
   scoped_ptr<content::WebContents> web_contents_;
   content::NotificationRegistrar registrar_;
   bool started_;
@@ -125,8 +130,12 @@ class DriveWebContentsManager : public
content::WebContentsObserver,
 
 DriveWebContentsManager::DriveWebContentsManager(
     Profile* profile,
+    const std::string& app_id,
+    const std::string& endpoint_url,
     const CompletionCallback& completion_callback)
     : profile_(profile),
+      app_id_(app_id),
+      endpoint_url_(endpoint_url),
       started_(false),
       completion_callback_(completion_callback),
       weak_ptr_factory_(this) {
@@ -140,12 +149,14 @@ DriveWebContentsManager::~DriveWebContentsManager() {
 
 void DriveWebContentsManager::StartLoad() {
   started_ = true;
-  const GURL url(kDriveOfflineEndpointUrl);
+  const GURL url(endpoint_url_);
   content::WebContents::CreateParams create_params(
         profile_, content::SiteInstance::CreateForURL(profile_, url));
 
   web_contents_.reset(content::WebContents::Create(create_params));
   web_contents_->SetDelegate(this);
+  extensions::ExtensionWebContentsObserver::CreateForWebContents(
+      web_contents_.get());
 
   content::NavigationController::LoadURLParams load_params(url);
   load_params.transition_type = content::PAGE_TRANSITION_GENERATED;
@@ -183,7 +194,10 @@ void DriveWebContentsManager::DidFailProvisionalLoad(
     int error_code,
     const string16& error_description,
     content::RenderViewHost* render_view_host) {
-  OnOfflineInit(false);
+  if (is_main_frame) {
+    LOG(WARNING) << "Failed to load WebContents to enable offline mode.";
+    OnOfflineInit(false);
+  }
 }
 
 void DriveWebContentsManager::DidFailLoad(
@@ -193,7 +207,10 @@ void DriveWebContentsManager::DidFailLoad(
     int error_code,
     const string16& error_description,
     content::RenderViewHost* render_view_host) {
-  OnOfflineInit(false);
+  if (is_main_frame) {
+    LOG(WARNING) << "Failed to load WebContents to enable offline mode.";
+    OnOfflineInit(false);
+  }
 }
 
 bool DriveWebContentsManager::ShouldCreateWebContents(
@@ -213,7 +230,7 @@ bool DriveWebContentsManager::ShouldCreateWebContents(
       extensions::ExtensionSystem::Get(profile_)->extension_service();
   const extensions::Extension *extension =
       service->GetInstalledApp(target_url);
-  if (!extension || extension->id() != kDriveHostedAppId)
+  if (!extension || extension->id() != app_id_)
     return true;
 
   // The background contents creation is normally done in Browser, but
@@ -223,7 +240,7 @@ bool DriveWebContentsManager::ShouldCreateWebContents(
 
   // Prevent redirection if background contents already exists.
   if (background_contents_service->GetAppBackgroundContents(
-      UTF8ToUTF16(kDriveHostedAppId))) {
+      UTF8ToUTF16(app_id_))) {
     return false;
   }
   BackgroundContents* contents = background_contents_service
@@ -231,7 +248,7 @@ bool DriveWebContentsManager::ShouldCreateWebContents(
                                  route_id,
                                  profile_,
                                  frame_name,
-                                 ASCIIToUTF16(kDriveHostedAppId),
+                                 ASCIIToUTF16(app_id_),
                                  partition_id,
                                  session_storage_namespace);
 
@@ -253,7 +270,7 @@ void DriveWebContentsManager::Observe(
     const std::string app_id = UTF16ToUTF8(
         content::Details<BackgroundContentsOpenedDetails>(details)
             ->application_id);
-    if (app_id == kDriveHostedAppId)
+    if (app_id == app_id_)
       OnOfflineInit(true);
   }
 }
@@ -263,7 +280,11 @@ void DriveWebContentsManager::Observe(
 
 DriveFirstRunController::DriveFirstRunController()
     : profile_(ProfileManager::GetDefaultProfile()),
-      started_(false) {
+      started_(false),
+      initial_delay_secs_(kInitialDelaySeconds),
+      web_contents_timeout_secs_(kWebContentsTimeoutSeconds),
+      drive_offline_endpoint_url_(kDriveOfflineEndpointUrl),
+      drive_hosted_app_id_(kDriveHostedAppId) {
 }
 
 DriveFirstRunController::~DriveFirstRunController() {
@@ -274,9 +295,10 @@ void DriveFirstRunController::EnableOfflineMode() {
     started_ = true;
     initial_delay_timer_.Start(
       FROM_HERE,
-      base::TimeDelta::FromSeconds(kInitialDelaySeconds),
+      base::TimeDelta::FromSeconds(initial_delay_secs_),
       this,
       &DriveFirstRunController::EnableOfflineMode);
+    return;
   }
 
   if (!UserManager::Get()->IsLoggedInAsRegularUser()) {
@@ -288,7 +310,7 @@ void DriveFirstRunController::EnableOfflineMode() {
 
   ExtensionService* extension_service =
       extensions::ExtensionSystem::Get(profile_)->extension_service();
-  if (!extension_service->GetExtensionById(kDriveHostedAppId, false)) {
+  if (!extension_service->GetExtensionById(drive_hosted_app_id_, false)) {
     LOG(WARNING) << "Drive app is not installed.";
     OnOfflineInit(false);
     return;
@@ -297,7 +319,7 @@ void DriveFirstRunController::EnableOfflineMode() {
   BackgroundContentsService* background_contents_service =
       BackgroundContentsServiceFactory::GetForProfile(profile_);
   if (background_contents_service->GetAppBackgroundContents(
-      UTF8ToUTF16(kDriveHostedAppId))) {
+      UTF8ToUTF16(drive_hosted_app_id_))) {
     LOG(WARNING) << "Background page for Drive app already exists";
     OnOfflineInit(false);
     return;
@@ -305,18 +327,44 @@ void DriveFirstRunController::EnableOfflineMode() {
 
   web_contents_manager_.reset(new DriveWebContentsManager(
       profile_,
+      drive_hosted_app_id_,
+      drive_offline_endpoint_url_,
       base::Bind(&DriveFirstRunController::OnOfflineInit,
                  base::Unretained(this))));
   web_contents_manager_->StartLoad();
   web_contents_timer_.Start(
       FROM_HERE,
-      base::TimeDelta::FromSeconds(kWebContentsTimeoutSeconds),
+      base::TimeDelta::FromSeconds(web_contents_timeout_secs_),
       this,
       &DriveFirstRunController::OnWebContentsTimedOut);
 }
 
+void DriveFirstRunController::AddObserver(Observer* observer) {
+  observer_list_.AddObserver(observer);
+}
+
+void DriveFirstRunController::RemoveObserver(Observer* observer) {
+  observer_list_.RemoveObserver(observer);
+}
+
+void DriveFirstRunController::SetDelaysForTest(int initial_delay_secs,
+                                               int timeout_secs) {
+  DCHECK(!started_);
+  initial_delay_secs_ = initial_delay_secs;
+  web_contents_timeout_secs_ = timeout_secs;
+}
+
+void DriveFirstRunController::SetAppInfoForTest(
+    const std::string& app_id,
+    const std::string& endpoint_url) {
+  DCHECK(!started_);
+  drive_hosted_app_id_ = app_id;
+  drive_offline_endpoint_url_ = endpoint_url;
+}
+
 void DriveFirstRunController::OnWebContentsTimedOut() {
   LOG(WARNING) << "Timed out waiting for web contents to opt-in";
+  FOR_EACH_OBSERVER(Observer, observer_list_, OnTimedOut());
   OnOfflineInit(false);
 }
 
@@ -329,8 +377,11 @@ void DriveFirstRunController::CleanUp() {
 
 void DriveFirstRunController::OnOfflineInit(bool success) {
   DCHECK(content::BrowserThread::CurrentlyOn(content::BrowserThread::UI));
-  ash::Shell::GetInstance()->system_tray_notifier()
-      ->NotifyDriveOfflineEnabled();
+  if (success) {
+   …
(message too large)

[commit-bot: I haz the power, 2013-11-14 18:16:21]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/tengs@chromium.org/71013003/270001

[commit-bot: I haz the power, 2013-11-14 19:19:55]

Retried try job too often on linux_rel for step(s) nacl_integration
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=linux_rel&...

[commit-bot: I haz the power, 2013-11-14 19:41:53]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/tengs@chromium.org/71013003/270001

[commit-bot: I haz the power, 2013-11-14 22:31:51]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/tengs@chromium.org/71013003/270001

[commit-bot: I haz the power, 2013-11-15 02:24:02]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/tengs@chromium.org/71013003/270001

[commit-bot: I haz the power, 2013-11-15 04:19:27]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/tengs@chromium.org/71013003/270001

[commit-bot: I haz the power, 2013-11-15 07:43:23]

Retried try job too often on win_rel for step(s) nacl_integration
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=win_rel&nu...

[commit-bot: I haz the power, 2013-11-15 08:35:12]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/tengs@chromium.org/71013003/270001

[commit-bot: I haz the power, 2013-11-15 10:26:21]

Retried try job too often on win_rel for step(s) nacl_integration
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=win_rel&nu...

[commit-bot: I haz the power, 2013-11-15 19:11:44]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/tengs@chromium.org/71013003/270001

[commit-bot: I haz the power, 2013-11-15 22:23:13]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/tengs@chromium.org/71013003/270001

[commit-bot: I haz the power, 2013-11-16 01:02:34]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/tengs@chromium.org/71013003/270001

[commit-bot: I haz the power, 2013-11-18 08:52:41]

Change committed as 235674