Cast Streaming API end-to-end browser_test.

Cast Streaming API end-to-end browser_test.

Similar to the TabCaptureApi.EndToEnd test: An extension is run that generates a rotating cycle of colors and audio tones, and the result is captured and streamed to an in-process Cast receiver.  In addition:

1. Refactored existing code out of the cast_receiver_app into a shared InProcessReceiver class.
2. Added a convenient StandaloneCastEnvironment for test apps.
3. Got rid of the cast_receiver_app prompts since the defaults are used >99% of the time.  Using the --enable-prompts command line arg will show them again.
4. Minor clean-ups in files touched.

NOTE: The new EndToEnd test is being submitted as disabled while outstanding bugs are worked by the team.  Will re-enable once the implementation is sufficiently stable for the bots.

R=hubbe@chromium.org, kalman@chromium.org

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=255805

[miu, 2014-03-01 02:45:53]

hubbe: PTAL.  If there's something I can do here in the code factoring to help
your perf tests down-the-road, let me know.

[hubbe, 2014-03-04 22:42:26]

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/DEPS
File chrome/browser/extensions/DEPS (right):

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/DE...
chrome/browser/extensions/DEPS:20: specific_include_rules = {
I wonder if it would be better to just add this for the directory, since the
performance tests will need the same thing?

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
File chrome/browser/extensions/cast_streaming_apitest.cc (right):

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
chrome/browser/extensions/cast_streaming_apitest.cc:142: base::AutoLock
auto_lock(lock_);
What is this lock for? Doesn't all of these functions run on the same thread?

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
chrome/browser/extensions/cast_streaming_apitest.cc:176:
UpdateExponentialMovingAverage(                             \
Why do we need a moving average?
A moving average of a median seems like massive overkill..

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
chrome/browser/extensions/cast_streaming_apitest.cc:179: UPDATE_FOR_PLANE(y,
kYPlane);
I think an uint8 medians[3], a const int planes[] = { ... } and a loop would be
prettier...

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
chrome/browser/extensions/cast_streaming_apitest.cc:208: *average += weight *
sample_value - weight * (*average);
Shouldn't this be:
*average += weight * sample_value  + (1 - weight) * (*average);
?
Or am I missing the intention here?

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
chrome/browser/extensions/cast_streaming_apitest.cc:225: memmove(dest, src,
width);
Are you sure we're actually allowed to modify the frames that we get from the
cast library?

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
chrome/browser/extensions/cast_streaming_apitest.cc:268: // media/cast library
has achieved sufficient stability.
Maybe file a bug too?

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
chrome/browser/extensions/cast_streaming_apitest.cc:279: // sender runs on port
2346.
This is not a good way to do it.
We should create a socket, set SO_REUSADDR on it, and then bind it to port 0.
Then we check what port we got and close the socket.
The port will be a high number and reserved by the OS so nobody else will get
that port for two minutes unless they bind to it specifically. (Which they
shouldn't do.)

I don't suppose there is a helper function for this already somewhere?

https://codereview.chromium.org/184813009/diff/1/chrome/test/data/extensions/...
File chrome/test/data/extensions/api_test/cast_streaming/end_to_end_sender.js
(right):

https://codereview.chromium.org/184813009/diff/1/chrome/test/data/extensions/...
chrome/test/data/extensions/api_test/cast_streaming/end_to_end_sender.js:26: //
Important: Blink the testing message so that the capture pipeline will
Do we need this if we are also changing colors?

https://codereview.chromium.org/184813009/diff/1/media/cast/cast_environment.cc
File media/cast/cast_environment.cc (right):

https://codereview.chromium.org/184813009/diff/1/media/cast/cast_environment....
media/cast/cast_environment.cc:86: // A convenience macro to efficiently check
and create a missing task runner
I don't think this is a good idea.
I think it would be better to create a utility class that does this by owning
the task runners directly.

https://codereview.chromium.org/184813009/diff/1/media/cast/cast_environment.h
File media/cast/cast_environment.h (right):

https://codereview.chromium.org/184813009/diff/1/media/cast/cast_environment....
media/cast/cast_environment.h:48: CastEnvironment();  // No logging.
This is only used for testing.
Seems better to keep it out of the production code.

https://codereview.chromium.org/184813009/diff/1/media/cast/test/utility/in_p...
File media/cast/test/utility/in_process_receiver.cc (right):

https://codereview.chromium.org/184813009/diff/1/media/cast/test/utility/in_p...
media/cast/test/utility/in_process_receiver.cc:137: // TODO(miu): Fix audio
decoder so that it never drops a request for the next
File a bug for this.

https://codereview.chromium.org/184813009/diff/1/media/cast/test/utility/in_p...
File media/cast/test/utility/in_process_receiver.h (right):

https://codereview.chromium.org/184813009/diff/1/media/cast/test/utility/in_p...
media/cast/test/utility/in_process_receiver.h:50: CastEnvironment* cast_env()
const { return cast_environment_.get(); }
Maybe return a scoped_refptr?

https://codereview.chromium.org/184813009/diff/1/media/cast/test/utility/in_p...
media/cast/test/utility/in_process_receiver.h:58: static AudioReceiverConfig
GetDefaultAudioConfig();
I don't think these functions belong in this class.

[miu, 2014-03-06 06:09:15]

Addressed your comments, and also got the test working now (except audio is
currently broken in the library).  PTAL.

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/DEPS
File chrome/browser/extensions/DEPS (right):

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/DE...
chrome/browser/extensions/DEPS:20: specific_include_rules = {
Actually, we should have a cast_streaming directory for our files.  So, I made
that dir, moved cast_streaming_apitest.cc to there, and created a new DEPS file
to let all files in that dir depend on media/cast.  LGTY?

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
File chrome/browser/extensions/cast_streaming_apitest.cc (right):

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
chrome/browser/extensions/cast_streaming_apitest.cc:142: base::AutoLock
auto_lock(lock_);
On 2014/03/04 22:42:26, hubbe wrote:
> What is this lock for? Doesn't all of these functions run on the same thread?

No.  All CastEnvironment threads are different from the unit test thread.  The
unit test thread calls WaitForColorAndTone() and blocks until a match has been
established.

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
chrome/browser/extensions/cast_streaming_apitest.cc:176:
UpdateExponentialMovingAverage(                             \
On 2014/03/04 22:42:26, hubbe wrote:
> Why do we need a moving average?
> A moving average of a median seems like massive overkill..

Good point.  Done.

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
chrome/browser/extensions/cast_streaming_apitest.cc:179: UPDATE_FOR_PLANE(y,
kYPlane);
On 2014/03/04 22:42:26, hubbe wrote:
> I think an uint8 medians[3], a const int planes[] = { ... } and a loop would
be
> prettier...

Done.

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
chrome/browser/extensions/cast_streaming_apitest.cc:208: *average += weight *
sample_value - weight * (*average);
On 2014/03/04 22:42:26, hubbe wrote:
> Shouldn't this be:
> *average += weight * sample_value  + (1 - weight) * (*average);
> ?
> Or am I missing the intention here?

Note: I'm using "*average += ..." and not "*average = ..."
                          ^^                       ^^^

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
chrome/browser/extensions/cast_streaming_apitest.cc:225: memmove(dest, src,
width);
On 2014/03/04 22:42:26, hubbe wrote:
> Are you sure we're actually allowed to modify the frames that we get from the
> cast library?

Yes.  There is no other reader of the frame.

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
chrome/browser/extensions/cast_streaming_apitest.cc:268: // media/cast library
has achieved sufficient stability.
On 2014/03/04 22:42:26, hubbe wrote:
> Maybe file a bug too?

Done.

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
chrome/browser/extensions/cast_streaming_apitest.cc:279: // sender runs on port
2346.
On 2014/03/04 22:42:26, hubbe wrote:
> This is not a good way to do it.

Done.  I had looked several weeks ago for the helper function, but found
nothing.  I've gone with your suggested "algorithm" to let the kernel figure out
the port for me.  I hope this works on Windows too.

https://codereview.chromium.org/184813009/diff/1/chrome/test/data/extensions/...
File chrome/test/data/extensions/api_test/cast_streaming/end_to_end_sender.js
(right):

https://codereview.chromium.org/184813009/diff/1/chrome/test/data/extensions/...
chrome/test/data/extensions/api_test/cast_streaming/end_to_end_sender.js:26: //
Important: Blink the testing message so that the capture pipeline will
On 2014/03/04 22:42:26, hubbe wrote:
> Do we need this if we are also changing colors?

Yes.  It causes the pipeline to generate more frames, allowing the test to
complete sooner.  This is how we do it in TabCaptureApiTest.EndToEnd.

https://codereview.chromium.org/184813009/diff/1/media/cast/cast_environment.cc
File media/cast/cast_environment.cc (right):

https://codereview.chromium.org/184813009/diff/1/media/cast/cast_environment....
media/cast/cast_environment.cc:86: // A convenience macro to efficiently check
and create a missing task runner
On 2014/03/04 22:42:26, hubbe wrote:
> I don't think this is a good idea.
> I think it would be better to create a utility class that does this by owning
> the task runners directly.

Done.

https://codereview.chromium.org/184813009/diff/1/media/cast/cast_environment.h
File media/cast/cast_environment.h (right):

https://codereview.chromium.org/184813009/diff/1/media/cast/cast_environment....
media/cast/cast_environment.h:48: CastEnvironment();  // No logging.
On 2014/03/04 22:42:26, hubbe wrote:
> This is only used for testing.
> Seems better to keep it out of the production code.

Done.

https://codereview.chromium.org/184813009/diff/1/media/cast/test/utility/in_p...
File media/cast/test/utility/in_process_receiver.cc (right):

https://codereview.chromium.org/184813009/diff/1/media/cast/test/utility/in_p...
media/cast/test/utility/in_process_receiver.cc:137: // TODO(miu): Fix audio
decoder so that it never drops a request for the next
On 2014/03/04 22:42:26, hubbe wrote:
> File a bug for this.

Already is one open.  Added link.

https://codereview.chromium.org/184813009/diff/1/media/cast/test/utility/in_p...
File media/cast/test/utility/in_process_receiver.h (right):

https://codereview.chromium.org/184813009/diff/1/media/cast/test/utility/in_p...
media/cast/test/utility/in_process_receiver.h:50: CastEnvironment* cast_env()
const { return cast_environment_.get(); }
On 2014/03/04 22:42:26, hubbe wrote:
> Maybe return a scoped_refptr?

Done.

https://codereview.chromium.org/184813009/diff/1/media/cast/test/utility/in_p...
media/cast/test/utility/in_process_receiver.h:58: static AudioReceiverConfig
GetDefaultAudioConfig();
On 2014/03/04 22:42:26, hubbe wrote:
> I don't think these functions belong in this class.

Done.  Moved to a test utility module.

[hubbe, 2014-03-06 19:54:42]

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/DEPS
File chrome/browser/extensions/DEPS (right):

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/DE...
chrome/browser/extensions/DEPS:20: specific_include_rules = {
On 2014/03/06 06:09:15, miu wrote:
> Actually, we should have a cast_streaming directory for our files.  So, I made
> that dir, moved cast_streaming_apitest.cc to there, and created a new DEPS
file
> to let all files in that dir depend on media/cast.  LGTY?

Sounds great.

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
File chrome/browser/extensions/cast_streaming_apitest.cc (right):

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
chrome/browser/extensions/cast_streaming_apitest.cc:142: base::AutoLock
auto_lock(lock_);
On 2014/03/06 06:09:15, miu wrote:
> On 2014/03/04 22:42:26, hubbe wrote:
> > What is this lock for? Doesn't all of these functions run on the same
thread?
> 
> No.  All CastEnvironment threads are different from the unit test thread.  The
> unit test thread calls WaitForColorAndTone() and blocks until a match has been
> established.

I missed the exclamation mark in the DCHECK in WaitForColorAndTone()...

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
chrome/browser/extensions/cast_streaming_apitest.cc:208: *average += weight *
sample_value - weight * (*average);
On 2014/03/06 06:09:15, miu wrote:
> On 2014/03/04 22:42:26, hubbe wrote:
> > Shouldn't this be:
> > *average += weight * sample_value  + (1 - weight) * (*average);
> > ?
> > Or am I missing the intention here?
> 
> Note: I'm using "*average += ..." and not "*average = ..."
>                           ^^                       ^^^

Ah tricky. Seems harder to read that way though.

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
chrome/browser/extensions/cast_streaming_apitest.cc:225: memmove(dest, src,
width);
On 2014/03/06 06:09:15, miu wrote:
> On 2014/03/04 22:42:26, hubbe wrote:
> > Are you sure we're actually allowed to modify the frames that we get from
the
> > cast library?
> 
> Yes.  There is no other reader of the frame.

I was more concerned about the vp8 library using the frame to construct the next
frame. But perhaps the frame we get is a copy...

https://codereview.chromium.org/184813009/diff/1/chrome/test/data/extensions/...
File chrome/test/data/extensions/api_test/cast_streaming/end_to_end_sender.js
(right):

https://codereview.chromium.org/184813009/diff/1/chrome/test/data/extensions/...
chrome/test/data/extensions/api_test/cast_streaming/end_to_end_sender.js:26: //
Important: Blink the testing message so that the capture pipeline will
On 2014/03/06 06:09:15, miu wrote:
> On 2014/03/04 22:42:26, hubbe wrote:
> > Do we need this if we are also changing colors?
> 
> Yes.  It causes the pipeline to generate more frames, allowing the test to
> complete sooner.  This is how we do it in TabCaptureApiTest.EndToEnd.

I know, I just wanted to make sure we still need it.

https://codereview.chromium.org/184813009/diff/20001/chrome/browser/extension...
File chrome/browser/extensions/api/cast_streaming/cast_streaming_apitest.cc
(right):

https://codereview.chromium.org/184813009/diff/20001/chrome/browser/extension...
chrome/browser/extensions/api/cast_streaming/cast_streaming_apitest.cc:273:
LOG(WARNING) << "Skipping this test since OSMesa is too slow on the bots.  "
It's still too slow, even with a window size of 128x128?
If that's the case, perhaps we should leave the window at the default size?

https://codereview.chromium.org/184813009/diff/20001/chrome/test/data/extensi...
File chrome/test/data/extensions/api_test/cast_streaming/end_to_end_sender.js
(right):

https://codereview.chromium.org/184813009/diff/20001/chrome/test/data/extensi...
chrome/test/data/extensions/api_test/cast_streaming/end_to_end_sender.js:103: //
TODO(miu): Reduce to 128x128 once cast v2 stops hard-coding
File a bug.

https://codereview.chromium.org/184813009/diff/20001/media/cast/cast_environm...
File media/cast/cast_environment.cc (right):

https://codereview.chromium.org/184813009/diff/20001/media/cast/cast_environm...
media/cast/cast_environment.cc:95: return main_thread_proxy_ &&
Is this rebase stuff, or did you change this?
If you changed it: Why?

https://codereview.chromium.org/184813009/diff/20001/media/cast/test/utility/...
File media/cast/test/utility/in_process_receiver.h (right):

https://codereview.chromium.org/184813009/diff/20001/media/cast/test/utility/...
media/cast/test/utility/in_process_receiver.h:56: // with the guarantee that the
system will eventually shut down.
This raises some questions about when it is safe to destroy this class.
Seems like we either need some way to wait until everything is stopped, or we
need to guarantee that destroying this class is *always* safe.

If the cast receiver works anything like the cast sender, then it should always
be safe to destroy the class.

https://codereview.chromium.org/184813009/diff/20001/media/cast/test/utility/...
File media/cast/test/utility/standalone_cast_environment.cc (right):

https://codereview.chromium.org/184813009/diff/20001/media/cast/test/utility/...
media/cast/test/utility/standalone_cast_environment.cc:14: : CastEnvironment(
Just use the 7-argument constructor, no need to special-case this.

[miu, 2014-03-07 22:40:28]

hubbe: PTAL.

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
File chrome/browser/extensions/cast_streaming_apitest.cc (right):

https://codereview.chromium.org/184813009/diff/1/chrome/browser/extensions/ca...
chrome/browser/extensions/cast_streaming_apitest.cc:208: *average += weight *
sample_value - weight * (*average);
Changed.

https://codereview.chromium.org/184813009/diff/20001/chrome/browser/extension...
File chrome/browser/extensions/api/cast_streaming/cast_streaming_apitest.cc
(right):

https://codereview.chromium.org/184813009/diff/20001/chrome/browser/extension...
chrome/browser/extensions/api/cast_streaming/cast_streaming_apitest.cc:273:
LOG(WARNING) << "Skipping this test since OSMesa is too slow on the bots.  "
On 2014/03/06 19:54:43, hubbe wrote:
> It's still too slow, even with a window size of 128x128?
> If that's the case, perhaps we should leave the window at the default size?

Fixed resolution issues, now that I know about the extra set of video parameters
that needed to be set in end_to_end_sender.js.

https://codereview.chromium.org/184813009/diff/20001/media/cast/cast_environm...
File media/cast/cast_environment.cc (right):

https://codereview.chromium.org/184813009/diff/20001/media/cast/cast_environm...
media/cast/cast_environment.cc:95: return main_thread_proxy_ &&
On 2014/03/06 19:54:43, hubbe wrote:
> Is this rebase stuff, or did you change this?
> If you changed it: Why?

I changed it.  Three reasons:

1. The protected constructor leaves these as NULL, so only the subclasses have
knowledge of whether these will be set or not.

2. Crash avoidance.

3. Logic: The method is asking whether the current thread is the one specified
by |identifier|.  That can never be true if xxx_thread_proxy_ is NULL, right?

https://codereview.chromium.org/184813009/diff/20001/media/cast/test/utility/...
File media/cast/test/utility/in_process_receiver.h (right):

https://codereview.chromium.org/184813009/diff/20001/media/cast/test/utility/...
media/cast/test/utility/in_process_receiver.h:56: // with the guarantee that the
system will eventually shut down.
On 2014/03/06 19:54:43, hubbe wrote:
> This raises some questions about when it is safe to destroy this class.
> Seems like we either need some way to wait until everything is stopped, or we
> need to guarantee that destroying this class is *always* safe.
> 
> If the cast receiver works anything like the cast sender, then it should
always
> be safe to destroy the class.

Re-worked this as discussed.  InProcessReceiver is no longer ref-counted, and
runs entirely on the Cast MAIN thread once Start() is invoked.

https://codereview.chromium.org/184813009/diff/20001/media/cast/test/utility/...
File media/cast/test/utility/standalone_cast_environment.cc (right):

https://codereview.chromium.org/184813009/diff/20001/media/cast/test/utility/...
media/cast/test/utility/standalone_cast_environment.cc:14: : CastEnvironment(
On 2014/03/06 19:54:43, hubbe wrote:
> Just use the 7-argument constructor, no need to special-case this.

Done.

[miu, 2014-03-07 22:41:42]

kalman: Need OWNERS approval for adding cast_streaming dir under
src/chrome/browser/extensions/api.

[not at google - send to devlin, 2014-03-07 22:43:30]

lgtm

[hubbe, 2014-03-07 22:48:50]

LGTM + 2 nits

https://codereview.chromium.org/184813009/diff/60001/chrome/browser/extension...
File chrome/browser/extensions/api/cast_streaming/cast_streaming_apitest.cc
(right):

https://codereview.chromium.org/184813009/diff/60001/chrome/browser/extension...
chrome/browser/extensions/api/cast_streaming/cast_streaming_apitest.cc:122: //
TODO(miu): Check audio tone too, once audio is fixed in the library.
bug?

https://codereview.chromium.org/184813009/diff/60001/media/cast/test/utility/...
File media/cast/test/utility/standalone_cast_environment.cc (right):

https://codereview.chromium.org/184813009/diff/60001/media/cast/test/utility/...
media/cast/test/utility/standalone_cast_environment.cc:16: NULL,
That's not what I meant...
Why not just pass in the task runners here?

[miu, 2014-03-08 02:20:49]

https://codereview.chromium.org/184813009/diff/60001/chrome/browser/extension...
File chrome/browser/extensions/api/cast_streaming/cast_streaming_apitest.cc
(right):

https://codereview.chromium.org/184813009/diff/60001/chrome/browser/extension...
chrome/browser/extensions/api/cast_streaming/cast_streaming_apitest.cc:122: //
TODO(miu): Check audio tone too, once audio is fixed in the library.
On 2014/03/07 22:48:50, hubbe wrote:
> bug?

Done.

https://codereview.chromium.org/184813009/diff/60001/media/cast/test/utility/...
File media/cast/test/utility/standalone_cast_environment.cc (right):

https://codereview.chromium.org/184813009/diff/60001/media/cast/test/utility/...
media/cast/test/utility/standalone_cast_environment.cc:16: NULL,
On 2014/03/07 22:48:50, hubbe wrote:
> That's not what I meant...
> Why not just pass in the task runners here?

The base class must be constructed before the derived one.  The
SingleThreadTaskRunner instances are not available until the threads are created
and started below.

[miu, 2014-03-08 02:20:55]

The CQ bit was checked by miu@chromium.org

[commit-bot: I haz the power, 2014-03-08 11:15:24]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/miu@chromium.org/184813009/80001

[commit-bot: I haz the power, 2014-03-08 12:51:53]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-03-08 12:51:53]

Retried try job too often on linux_chromeos for step(s) app_list_unittests,
ash_unittests, aura_unittests, base_unittests, browser_tests,
cacheinvalidation_unittests, check_deps, check_deps2git, chromeos_unittests,
components_unittests, content_browsertests, content_unittests, crypto_unittests,
dbus_unittests, device_unittests, events_unittests, google_apis_unittests,
gpu_unittests, interactive_ui_tests, ipc_tests, jingle_unittests,
keyboard_unittests, media_unittests, net_unittests, ppapi_unittests,
printing_unittests, sandbox_linux_unittests, sql_unittests, sync_unit_tests,
unit_tests
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=linux_chro...

[miu, 2014-03-08 21:33:35]

The CQ bit was checked by miu@chromium.org

[commit-bot: I haz the power, 2014-03-08 21:33:42]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/miu@chromium.org/184813009/100001

[commit-bot: I haz the power, 2014-03-08 22:49:33]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-03-08 22:49:34]

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

[miu, 2014-03-08 23:30:13]

The CQ bit was checked by miu@chromium.org

[commit-bot: I haz the power, 2014-03-08 23:30:33]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/miu@chromium.org/184813009/100001

[commit-bot: I haz the power, 2014-03-09 00:27:16]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-03-09 00:27:17]

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

[miu, 2014-03-09 01:42:02]

Committed patchset #6 manually as r255805 (presubmit successful).