Add UMA stats for open audio input device failure.

Add UMA stats for open audio input device failure.

* Change UMA stats for capture startup to an enum. It retains the meanings of the current (bool) values 0 and 1.
* Add values for fail to create and open stream (device).
* Remove |enable_nodata_timer| which doesn't have any use.

BUG=399835
Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=289288

[Henrik Grunell, 2014-08-12 12:23:43]

Henrik: media/audio/*
Ilya: tools/metrics/histograms/histograms.xml

[henrika (OOO until Aug 14), 2014-08-12 12:29:35]

LGTM. Thanks!

[Ilya Sherman, 2014-08-12 17:30:48]

LGTM % nits:

https://codereview.chromium.org/460993002/diff/1/media/audio/audio_input_cont...
File media/audio/audio_input_controller.cc (right):

https://codereview.chromium.org/460993002/diff/1/media/audio/audio_input_cont...
media/audio/audio_input_controller.cc:54: // permitted operation is to add new
elements before NUM_STREAM_OPEN_RESULT.
nit: Please update the name "NUM_STREAM_OPEN_RESULT" to
"CAPTURE_STARTUP_RESULT_MAX".

https://codereview.chromium.org/460993002/diff/1/media/audio/audio_input_cont...
media/audio/audio_input_controller.cc:60: CAPTURE_STARTUP_RESULT_MAX =
CAPTURE_STARTUP_OPEN_STREAM_FAILED
Optional nit: It's a little cleaner, IMO, to have this enum member be
CAPTURE_STARTUP_RESULT_COUNT, and have that be equal to 4.  That way you can
avoid the "+ 1" below; and also anyone adding values to this enum doesn't need
to remember to update this line as well.  For that second reason, I'd also omit
the explicit value assignments for all of the enum members -- the compiler is
better at assigning values in a linear sequence than humans are.

[Henrik Grunell, 2014-08-13 12:31:30]

Code review fix. Also changed the name of a function.

https://codereview.chromium.org/460993002/diff/1/media/audio/audio_input_cont...
File media/audio/audio_input_controller.cc (right):

https://codereview.chromium.org/460993002/diff/1/media/audio/audio_input_cont...
media/audio/audio_input_controller.cc:54: // permitted operation is to add new
elements before NUM_STREAM_OPEN_RESULT.
On 2014/08/12 17:30:48, Ilya Sherman wrote:
> nit: Please update the name "NUM_STREAM_OPEN_RESULT" to
> "CAPTURE_STARTUP_RESULT_MAX".

Done.

https://codereview.chromium.org/460993002/diff/1/media/audio/audio_input_cont...
media/audio/audio_input_controller.cc:60: CAPTURE_STARTUP_RESULT_MAX =
CAPTURE_STARTUP_OPEN_STREAM_FAILED
On 2014/08/12 17:30:48, Ilya Sherman wrote:
> Optional nit: It's a little cleaner, IMO, to have this enum member be
> CAPTURE_STARTUP_RESULT_COUNT, and have that be equal to 4.  That way you can
> avoid the "+ 1" below; and also anyone adding values to this enum doesn't need
> to remember to update this line as well.  For that second reason, I'd also
omit

The presubmit check actually complained about that, it had to be done this way.
It seems it could be specific for //media.

> the explicit value assignments for all of the enum members -- the compiler is
> better at assigning values in a linear sequence than humans are.

I think it's easier to see the match with the enum in histograms.xml when
written explicitly.

[Henrik Grunell, 2014-08-13 12:31:40]

The CQ bit was checked by grunell@chromium.org

[commit-bot: I haz the power, 2014-08-13 12:32:55]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/grunell@chromium.org/460993002/20001

[commit-bot: I haz the power, 2014-08-13 14:08:22]

FYI, CQ is re-trying this CL (attempt #1).
The failing builders are:
  mac_gpu_retina_triggered_tests on tryserver.chromium.gpu
(http://build.chromium.org/p/tryserver.chromium.gpu/builders/mac_gpu_retina_tr...)

[commit-bot: I haz the power, 2014-08-13 14:30:22]

Committed patchset #2 (20001) as 289288