MediaStreamTrack.getSources list is now updated when input devices is disabled on Windows

MediaStreamTrack.getSources list is now updated when input devices is disabled on Windows.

BUG=395733
TEST=https://simpl.info/getusermedia/sources/

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=289019

[henrika (OOO until Aug 14), 2014-08-12 11:47:19]

I found that the reported issue was due to a too narrow filter in
AudioDeviceListenerWin. It did not allow Disabled events to pass which resulted
in a failure to update the device list for  that case.

I found that we are OK without this filter.

Tried many different cases: disable/enable device, remove/insert device (being
disabled and enabled). All seems to work now.

[henrika (OOO until Aug 14), 2014-08-12 11:50:36]

To summarize the new state:

- Run https://simpl.info/getusermedia/sources/ with all devices enabled.
- Disable device X
- Refresh page => device X is now removed from list.
- Enable device X
- Refresh page => device X is now added to the list.

Same for remove/insert device X.

[xians, 2014-08-12 11:59:09]

https://codereview.chromium.org/459363003/diff/1/content/browser/renderer_hos...
File content/browser/renderer_host/media/media_stream_manager.cc (right):

https://codereview.chromium.org/459363003/diff/1/content/browser/renderer_hos...
content/browser/renderer_host/media/media_stream_manager.cc:1989: DVLOG(1) <<
"OnDevicesChanged(DEVTYPE_AUDIO_CAPTURE)";
do you still want these debug loggings? If not, please remove.

https://codereview.chromium.org/459363003/diff/1/media/audio/win/audio_device...
File media/audio/win/audio_device_listener_win.cc (right):

https://codereview.chromium.org/459363003/diff/1/media/audio/win/audio_device...
media/audio/win/audio_device_listener_win.cc:128: std::string
new_state_str("UNDEFINED");
Have you verified that you only get one event when you plugin or unplug a
device?

https://codereview.chromium.org/459363003/diff/1/media/audio/win/audio_device...
media/audio/win/audio_device_listener_win.cc:138: DVLOG(1) <<
"AudioDeviceListenerWin::OnDeviceStateChanged () "
I am not sure if this logging adds any value after you verify the behaviours.

[henrika (OOO until Aug 14), 2014-08-12 12:04:39]

https://codereview.chromium.org/459363003/diff/1/content/browser/renderer_hos...
File content/browser/renderer_host/media/media_stream_manager.cc (right):

https://codereview.chromium.org/459363003/diff/1/content/browser/renderer_hos...
content/browser/renderer_host/media/media_stream_manager.cc:1989: DVLOG(1) <<
"OnDevicesChanged(DEVTYPE_AUDIO_CAPTURE)";
I find them useful and would like to keep.

https://codereview.chromium.org/459363003/diff/1/media/audio/win/audio_device...
File media/audio/win/audio_device_listener_win.cc (right):

https://codereview.chromium.org/459363003/diff/1/media/audio/win/audio_device...
media/audio/win/audio_device_listener_win.cc:128: std::string
new_state_str("UNDEFINED");
I do get two callbacks but with the same new_state, hence the old filter did not
make a difference.

https://codereview.chromium.org/459363003/diff/1/media/audio/win/audio_device...
media/audio/win/audio_device_listener_win.cc:138: DVLOG(1) <<
"AudioDeviceListenerWin::OnDeviceStateChanged () "
It would have done this debugging more easy if it had existed... I find it
useful to track down corner cases.

[xians, 2014-08-12 12:15:55]

On 2014/08/12 12:04:39, henrika wrote:
>
https://codereview.chromium.org/459363003/diff/1/content/browser/renderer_hos...
> File content/browser/renderer_host/media/media_stream_manager.cc (right):
> 
>
https://codereview.chromium.org/459363003/diff/1/content/browser/renderer_hos...
> content/browser/renderer_host/media/media_stream_manager.cc:1989: DVLOG(1) <<
> "OnDevicesChanged(DEVTYPE_AUDIO_CAPTURE)";
> I find them useful and would like to keep.
> 
>
https://codereview.chromium.org/459363003/diff/1/media/audio/win/audio_device...
> File media/audio/win/audio_device_listener_win.cc (right):
> 
>
https://codereview.chromium.org/459363003/diff/1/media/audio/win/audio_device...
> media/audio/win/audio_device_listener_win.cc:128: std::string
> new_state_str("UNDEFINED");
> I do get two callbacks but with the same new_state, hence the old filter did
not
> make a difference.
> 
>
https://codereview.chromium.org/459363003/diff/1/media/audio/win/audio_device...
> media/audio/win/audio_device_listener_win.cc:138: DVLOG(1) <<
> "AudioDeviceListenerWin::OnDeviceStateChanged () "
> It would have done this debugging more easy if it had existed... I find it
> useful to track down corner cases.

OK, thanks for the clarification, lgtm

[henrika (OOO until Aug 14), 2014-08-12 12:16:53]

Tommi: care to take a look as well?

[tommi (sloooow) - chröme, 2014-08-12 13:00:00]

Good work on fixing!  I'd like the cl to be just the two lines that got removed
(remove the logging code) and assuming that, lgtm.

[henrika (OOO until Aug 14), 2014-08-12 13:40:05]

Thanks Tommi, done ;-)

[henrika (OOO until Aug 14), 2014-08-12 13:41:57]

The CQ bit was checked by henrika@chromium.org

[commit-bot: I haz the power, 2014-08-12 13:42:50]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/henrika@chromium.org/459363003/40001

[commit-bot: I haz the power, 2014-08-12 18:03:30]

Change committed as 289019