Unit tests:
* Add net_test_stack_rfcomm
* Use a compiler trick to compile production RFCOMM code against L2CAP
  and BTM header, but fake L2CAP and BTM source code so that we can stub
  statically defined L2CAP and BTM code using mocks
* Rename net_test_rfcomm to net_test_rfcomm_suite
* Add mocked L2CAP and BTM layer to allow packet replay from L2CAP layer
  to RFCOMM layer in order to reproduce time dependent issues such as
  connection collision
* Add a number of utility methods that generates RFCOMM packet bytes
  using supported parameters
* Add unit test for above utility methods
* Add suite of unit tests named StackRfcommTest, including:
 - SingleServerConnectionHelloWorld
 - MultiServerPortSameDeviceHelloWorld
 - SameServerPortMultiDeviceHelloWorld
 - SingleClientConnectionHelloWorld
 - SameClientPortMultiDeviceHelloWorld
* These tests supply L2CAP packets and API calls to trigger RFCOMM
  connection setup in various scenarios
* Added logging method to output debug log via VLOG(1) logging level and
  allow compile time configuration of these logs through a constant flag

Bug: 77224743
Test: ./test/run_unit_tests.sh net_test_stack_rfcomm
      testplans/details/158641/3975
Change-Id: I9d59843603cd36394c3736670bcf3c39dea26674

Bug: 63932139
Bug: 76201991
Test: Manual A2DP testing (A2DP offload enabled and disabled)
Change-Id: Idcc2af40bde09758e060861ac518c06b7551212a

Bug: 77281039
Test: Manual - connect Headset and switch codecs while playing audio
Change-Id: Icb89b0d78195fd654d66bc1334b121cf62c55be1

When writing UUID into stream with gatt_build_uuid_to_stream, we always
use 16bit or 128bit representation. In GATT code returning service UUID,
we were returning 32bit representation lenght, even though 128bit
representation was being returned in the stream. This resulted in
invalid GATT database content being returned, and the service not being
visible.

Bug: 66912853
Test: create GATT database with 32bit UUID, discover from remote device
Change-Id: I791a518ab080bd99db0a8be18d97e865c838fc7e

* changes:
  Include only the Basic Capability for AVDTP GetCapabilities response
  Handle properly AVDTP SetConfig from the A2DP Sink device

Some HF android-based devices need to disable
wide band speech due to interop issues.

Test: Pair and test phone calls
Bug: 65053641

Change-Id: Ie8c4a50c40149562d6b52336fba1a61f1abf5198

Added new function bta_av_free_scb() to explicitly
free tBTA_AV_SCB struct.

Bug: 78034873
Test: Code compilation
Change-Id: Ic738c64f61fd273bc2988fa98cb7cedebf1c4b76

Bug: 78134184
Test: Run host native test net_test_avrcp
Change-Id: I57dd0040aa0e97f623cff97a28dc98e196414559

We will need characteristic definition handle, and handle of end of
included service in the future. Store it properly.

Bug: 67056747
Test: GattCacheTest.nv_attr_to_binary_test
Change-Id: I170e1a39d3c055efddd91639f756d74eb4ecc304

AVDTP GetCapabilities request should return only the Basic
Capabilities, as defined in the AVDTP 1.3 spec.
Otherwise, returning a capability such as Delay Reporting would
result in some remote devices droping the AVDTP connection

Also:
 * Print SEP protocol service capabilities per Stream Control
   Block when dumping the state to a bugreport
 * Add extra log messages

Bug: 77668253
Test: Manual: connect with Momentum M2 Headset and examine the
      AVDTP GetCapabilities response.

Change-Id: I34f2e8ff586115a38c3436b2b6300514f65346a9

If the remote device is proactive and sends AVDTP SetConfig after
re-connection before the local device gets the chance to do it, the
internal codec setup state might be inconsistent.

* Fix the internal logic when the local device is Acceptor inside
  file bta_av_co.cc, and simplify some of the code.
* Fix the handling of p_scb->sep_info_idx inside bta_av_save_caps()
  when receiving capabilities from the remote device.
* Add new A2dpCodecConfig::setPeerCodecCapabilities() method that is
  implemented by each codec.

Bug: 77525584
Test: Manual - initiate connection by Momentum 2.0 Wireless Headset.
      Connect/disconnect/reconnect multiple headsets.

Change-Id: I456df7c8a2fa0758c0908a4628c4dfae3259dbdb

ESR06 (errata 4065) added sentence:
"The End Group Handle of the last service in a device can be 0xFFFF."
It talks about last service, not last primary service. Right now,
secondary service added after last primary service will result in in
last primary service being corrupted - descriptor added to last
characteristic, and some characteristics added.

Bug: 78107486
Change-Id: I6a146fce2c946cdff176ff79f147830e29d3b807

Bug: 77289942
Test: Manual
Change-Id: I5b7ca6e485788d69f574bb74d2ef423507b183ff

Before, the upper layers would be informed when a device connected that
supported absolute volume. Now even devices that don't support absolute
volume get reported.

Bug: 33962015
Test: Run host native test net-test-avrcp
Change-Id: If083ccbb9ba9a170c5bca359e632585d826c6fa5

Use LDAC ABR as default option in A2DP Offload mode.

Bug: 77879487
Test: Manual
Change-Id: Iec76691f8a83cd47aa7bf38efc6554139a59bb89

If host gets link_key_request event from the controller, it means peer
device started the authentication.
In this case, host should not start the authentication because it will
be caused the authentication collision by controller side.

This sequence can be referred at BT 4.2 core specification document,
p.1388, Figure 4.2.

Bug: 74241491
Test: Manual
      - Connect Android wear to Android phone
      - Turn off BT on Android wear
      - Turn on BT on Android wear
      - Observe the logcat whether connection time is within 30 seconds
      Android Wear Stability Test

Change-Id: I138d45316ea75dd96444c36bde36ce15a64b669a

Add new test connection event (ce) parameters, min_ce and max_ce, to
the SL4A tests for LE Connection-oriented Channel (CoC) feature.

Test: Run cmd: act.py -c $MY_SL4A_CONFIG -tc BleCoc2ConnTest
Bug: 77528723

Change-Id: Ifea040f0eb105f89539e6b74f5c2edc76147a065

AVRCP browsing was unstable on the sink side due to bytes
being used instead of integers for get_folder_items_cmd
and a lack of deep copy in the data buffers.  This patch
allows both browse and vendor commands to work correctly.

Bug: 72496280
Test: stream music and browse at the same time.
Change-Id: Id4d34101c918947450923b68e087e6851ce99130

Item A: Add messages to log profile connections
* Add ProfileUsageStats message in bluetooth.proto
* Add BluetoothProfileId enum to represent each profile
* Add num_times_connected item to log how many times a profile is
  connected

Item B: Add messages to log headset profile connections for HSP and HFP
* Add HeadsetProfileConnectionStats message in bluetooth.proto
* Add HeadsetProfileType enum to represent each headset profile type
* Add num_times_connected item to log how many times each headset
  profile is connected

* Add unit tests in metrics_tests
* Remove "clear" flag in various metrics dumping methods to make sure
  that we clean up metrics every time we dump

Bug: 77476285
Test: make, net_test_osi, BtFunhausMetricsTest,
      adb shell dumpsys bluetooth_manager --proto-bin
Change-Id: Ib18948c50ddb98ab7472f7b51a9bb98d153071c7
Merged-In: Ib18948c50ddb98ab7472f7b51a9bb98d153071c7

Bug: 72817086
Test: make, no functional change
Change-Id: I0df4efa62de00956eea1b5b5d244f4821bcaf6fc

AVRC_Close needs to be called whenever a connection is closed otherwise
the AVCT control block is never cleaned up.

Bug: 77633048
Test: Run host native test net_test_avrcp and connect/disconnect from a
device multiple times.

Change-Id: Ia6a1b2606a87803316aad41ae27db858bf922548

Rename A2DP_DumpCodecInfo() to A2DP_CodecInfoString() and change
the usage. Previously, A2DP_DumpCodecInfo() would print the
the codec information only if LOG_VERBOSE() was enabled in compile time.
The new A2DP_CodecInfoString() now returns human-readable std::string
with the codec information.
That string can be used in debug log messages that can be enabled
in run-time.

Bug: 77525584
Test: Manual - Examine log messages when A2DP streaming with and
      without debug log messages enabled.

Change-Id: Idd440d6c9e908520132feeeb5388d3e6aefa26db

Bug: 77874113
Test: Check snoop logs to see that the broadcasted AVRCP version is 1.4
Change-Id: If36bccc58d04dffe46286ee8d18785b680b9de8a

It is already possible to set server connection sniff timeouts.

Bug: 70309445
Test: inspect timeouts via logs
Change-Id: Ibc19dd4f77aa28388169b2ce70d12200020e6a4d

* Report AT+BIA command, AG indicator activation command, to upper layer
  so that upper layer can decide on whether or not to subscribe to these
  indicators

Bug: 77655992
Test: make

Change-Id: Ic330d6784c5edf20683cbfedda3fa54525878ce8

Bug: 76456293
Test: Listen to music on a device that supports delay reporting
Change-Id: I4ca1c31819594750a7beb1d11ca5b8d64a7f836c

Bug: 75290651
Test: Manual

Change-Id: I30bda23667ac911bf21da5705d248833f919f8bf
Signed-off-by: Dmitry Shmidt <dimitrysh@google.com>

Instead of trying to re-register for a rejected volume changed
notification, disable absolute volume. This prevents spinning if a
remote device continuously rejects all attempts to register. A volume
level of -2 will be used to represent that the volume notification was
rejected.

Bug: 77238060
Test: Run host native test net_test_avrcp
Change-Id: I228524fb30348ca691d0792f0c7bcc4653d1fcef