Fix discovery of Smart Lock service in Proximity Auth over Blutooth Low Energy.

Fix discovery of Smart Lock service in Proximity Auth over Blutooth Low Energy.

This CL adds observers Bluetooth device changed and device removed
notifications. Each time a device changes, it checks if the device advertises
the Smart Lock service UUID. If so and if there is no pending connection to
that device, then it tries to connect to it.

This CL also forgets the Bluetooth device that advertises Smart Lock. This
ensures that the device is disconnected and its entry removed fron the Bluetooth
device controller. Note that this strategy is fine as Bluetooth Low Energy
device that advertises Smart Lock service UUID is only used by the Proximity
Auth component.

TEST=On Chrome OS, turn on Smart Lock over Bluetooth Low Energy in settings.
Install a Smart Lock app on a BLE device. On Chrome OS, sign in and then lock
the screen. Start the Smart Lock app on a BLE device and observe that Chrome OS
magically unlocks.
BUG=479670
Committed: https://crrev.com/18d21e1eccf868ec06d59e65a93314148260e749
Cr-Commit-Position: refs/heads/master@{#327495}

[msarda, 2015-04-29 13:01:02]

msarda@chromium.org changed reviewers:
+ sacomoto@chromium.org

[msarda, 2015-04-29 13:01:03]

Please take a look.

[sacomoto, 2015-04-29 13:32:43]

LGTM.

https://codereview.chromium.org/1113793002/diff/20001/components/proximity_au...
File components/proximity_auth/ble/bluetooth_low_energy_connection_finder.cc
(right):

https://codereview.chromium.org/1113793002/diff/20001/components/proximity_au...
components/proximity_auth/ble/bluetooth_low_energy_connection_finder.cc:75:
BluetoothDevice* device) {
Nit: please put a DCHECK(device) here too.

https://codereview.chromium.org/1113793002/diff/20001/components/proximity_au...
components/proximity_auth/ble/bluetooth_low_energy_connection_finder.cc:85:
const auto& i = pending_connections_.find(device);
There is the (pretty rare) case when this function is called two times, and the
first call didn't add the device to pending_connection when the second call
looks for the device.

[msarda, 2015-04-29 14:11:33]

https://codereview.chromium.org/1113793002/diff/20001/components/proximity_au...
File components/proximity_auth/ble/bluetooth_low_energy_connection_finder.cc
(right):

https://codereview.chromium.org/1113793002/diff/20001/components/proximity_au...
components/proximity_auth/ble/bluetooth_low_energy_connection_finder.cc:75:
BluetoothDevice* device) {
On 2015/04/29 13:32:43, sacomoto wrote:
> Nit: please put a DCHECK(device) here too.

Done.

https://codereview.chromium.org/1113793002/diff/20001/components/proximity_au...
components/proximity_auth/ble/bluetooth_low_energy_connection_finder.cc:85:
const auto& i = pending_connections_.find(device);
On 2015/04/29 13:32:43, sacomoto wrote:
> There is the (pretty rare) case when this function is called two times, and
the
> first call didn't add the device to pending_connection when the second call
> looks for the device. 

This code is single threaded so there is no race condition here.

[msarda, 2015-04-29 14:11:36]

The CQ bit was checked by msarda@chromium.org

[msarda, 2015-04-29 14:11:36]

The patchset sent to the CQ was uploaded after l-g-t-m from
sacomoto@chromium.org
Link to the patchset: https://codereview.chromium.org/1113793002/#ps60001
(title: "Add DCHECKs")

[commit-bot: I haz the power, 2015-04-29 14:12:16]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1113793002/60001

[commit-bot: I haz the power, 2015-04-29 14:12:17]

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

[commit-bot: I haz the power, 2015-04-29 14:12:17]

No LGTM from a valid reviewer yet. Only full committers are accepted.
Even if an LGTM may have been provided, it was from a non-committer,
_not_ a full super star committer.
See http://www.chromium.org/getting-involved/become-a-committer
Note that this has nothing to do with OWNERS files.

[droger, 2015-04-29 14:41:03]

droger@chromium.org changed reviewers:
+ droger@chromium.org

[droger, 2015-04-29 14:41:03]

lgtm

https://codereview.chromium.org/1113793002/diff/60001/components/proximity_au...
File components/proximity_auth/ble/bluetooth_low_energy_connection_finder.cc
(right):

https://codereview.chromium.org/1113793002/diff/60001/components/proximity_au...
components/proximity_auth/ble/bluetooth_low_energy_connection_finder.cc:26: void
EmptyErrorCallback() {
Nit: you can use DoNothing() in base/bind_helpers.h

https://codereview.chromium.org/1113793002/diff/60001/components/proximity_au...
components/proximity_auth/ble/bluetooth_low_energy_connection_finder.cc:70:
VLOG(1) << "Device added: " << device->GetAddress();
Maybe VLOG is fine here, but have you considered using DLOG instead?

https://codereview.chromium.org/1113793002/diff/60001/components/proximity_au...
File components/proximity_auth/ble/bluetooth_low_energy_connection_finder.h
(right):

https://codereview.chromium.org/1113793002/diff/60001/components/proximity_au...
components/proximity_auth/ble/bluetooth_low_energy_connection_finder.h:102: //
True if a connection was established to a remove device that has the
s/remove/remote/ ?

[droger, 2015-04-29 14:42:01]

https://codereview.chromium.org/1113793002/diff/60001/components/proximity_au...
File components/proximity_auth/ble/bluetooth_low_energy_connection_finder.cc
(right):

https://codereview.chromium.org/1113793002/diff/60001/components/proximity_au...
components/proximity_auth/ble/bluetooth_low_energy_connection_finder.cc:70:
VLOG(1) << "Device added: " << device->GetAddress();
On 2015/04/29 14:41:03, droger wrote:
> Maybe VLOG is fine here, but have you considered using DLOG instead?

I meant DVLOG.

[msarda, 2015-04-29 14:45:12]

https://codereview.chromium.org/1113793002/diff/60001/components/proximity_au...
File components/proximity_auth/ble/bluetooth_low_energy_connection_finder.cc
(right):

https://codereview.chromium.org/1113793002/diff/60001/components/proximity_au...
components/proximity_auth/ble/bluetooth_low_energy_connection_finder.cc:26: void
EmptyErrorCallback() {
On 2015/04/29 14:41:03, droger wrote:
> Nit: you can use DoNothing() in base/bind_helpers.h

Done.

https://codereview.chromium.org/1113793002/diff/60001/components/proximity_au...
components/proximity_auth/ble/bluetooth_low_energy_connection_finder.cc:70:
VLOG(1) << "Device added: " << device->GetAddress();
On 2015/04/29 14:41:03, droger wrote:
> Maybe VLOG is fine here, but have you considered using DLOG instead?

We want to use VLOG to be able to debug on Chromebook when deploying Chrome (not
Chromium). As this code is only run on the screenlock, I do not see any
performance issue here.

https://codereview.chromium.org/1113793002/diff/60001/components/proximity_au...
File components/proximity_auth/ble/bluetooth_low_energy_connection_finder.h
(right):

https://codereview.chromium.org/1113793002/diff/60001/components/proximity_au...
components/proximity_auth/ble/bluetooth_low_energy_connection_finder.h:102: //
True if a connection was established to a remove device that has the
On 2015/04/29 14:41:03, droger wrote:
> s/remove/remote/ ?

Done.

[msarda, 2015-04-29 14:45:15]

The CQ bit was checked by msarda@chromium.org

[msarda, 2015-04-29 14:45:16]

The patchset sent to the CQ was uploaded after l-g-t-m from droger@chromium.org,
sacomoto@chromium.org
Link to the patchset: https://codereview.chromium.org/1113793002/#ps80001
(title: "Nit")

[commit-bot: I haz the power, 2015-04-29 14:45:35]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1113793002/80001

[commit-bot: I haz the power, 2015-04-29 15:38:52]

Committed patchset #5 (id:80001)

[commit-bot: I haz the power, 2015-04-29 15:40:30]

Patchset 5 (id:??) landed as
https://crrev.com/18d21e1eccf868ec06d59e65a93314148260e749
Cr-Commit-Position: refs/heads/master@{#327495}