Some players will send a null object as parameter for
playstate change callback, without updating the
MediaSession. These events should still trigger an update
by using the Mediasession metadata if it is different.

Bug: 297191838
Tag: #feature
Test: atest BluetoothInstrumentationTests
Change-Id: Ic5a5b6a68e616af939a4e42d7d64476932dc38d3

* changes:
  BQR: Fix event len for v5.0 incompatible devices
  Revert "BQR: Log error instead of crash if event too short"

Bug: 292162606
Test: atest BluetoothOppBtEnableActivityTest
Change-Id: Ie128f73dba172aaec2f9bd1800922fafcabaef61

Resolved 'Missing Library' error for android.hardware.bluetooth.audio-V2-ndk.so

Test: ./btcore_property_fuzzer
Bug: 216476246

Ignore-AOSP-First: Code of git_master is ahead of AOSP master
(cherry picked from https://googleplex-android-review.googlesource.com/q/commit:295c1b256a2717151574e24d05377ebf0a6193bb)
Merged-In: I8485a6ce391a021975fa6123e4d5932ba9aa8f5d
Change-Id: I8485a6ce391a021975fa6123e4d5932ba9aa8f5d

Bug: 293849013
Test: m libbtif-core
Test: manual - check BQR event logs

Change-Id: Ic9bced9c1f19527c586d9f62ee02fcda8e294fc8

Bug: 292010892
Tag: #bug
Test: atest bluetooth_test_broadcaster bluetooth_test_broadcaster_state_machine
Change-Id: I0e09aa8d22b0e94c60426052e50d09ee739dbc5d

Bug: 293849013
Test: m libbtif-core
Test: manual - check BQR event logs

This reverts commit f809cb86.

Change-Id: I7f572aabb7f201d7e7ca84a082346cc0886959c7

This imports log_panics to print log when panic! is reached.
An example of the log is:
thread 'main' panicked at 'xxx': system/gd/rust/linux/stack/src/
bluetooth_admin.rs:118

Bug: 295133385
Test: mma -j32
Test: locally modify the code to trigger panic and verify the log is in
/var/log/bluetooth.log
Tag: #floss

Change-Id: I217ccd30542c5e429abfcd58626a251307353049

Test: mmm packages/modules/Bluetooth
Bug: 263885771
Change-Id: Ie39288c2d7a0b607c5ba2896e4d37812bc3ea0b1

* changes:
  le_audio: Introduce Unicast fallback after Broadcast stop
  broadcast: Introduce pending broadcast
  broadcast: Introduce broadcasts descriptor

* changes:
  Remove allocation_tracker
  Inline AlarmTestHarness in its only user

ACL packets always contain L2CAP packets, so the assembler
uses the size of the L2CAP packet to determine when a
packet is completed.  If packets are smaller than the size
of an L2CAP header, wait to see if more will come before
discarding them.

Bug: 292933138
Test: atest bluetooth_test_gd_unit
Change-Id: I010f9c8b31eb4d2298cc53c9ff0a9bd6b180dc01

Bug: 292010892
Tag: #bug
Test: atest bluetooth_test_gd_unit
Test: manual check LEA broadcast not auto rotate nrpa
Change-Id: I042cc8665322d1732e2f68e2c47630238c30dc57

* changes:
  leaudio: Improve rejecting stream from remote device
  leaudio: Do not start stream if context not available but supported
  leaudio: Make sure context and locations are updated correctly
  leaudio: Remove not needed operation and return parameter
  leaudio: Add helper functions to inject context types

Those two defaults are providing the same properties

Test: mmm packages/modules/Bluetooth
Bug: 279502784
Change-Id: I62ada5912eeee005b54c9009ef7fc4da28150aa1

Bug: 279502784
Change-Id: I44daefae142ddec521a4e245e57d6ab1e706e7fe
Test: mmm packages/modules/Bluetooth

Bug: 279502784
Change-Id: I8b6400eb70b5078287f4d2d6cff5b71010b42559
Test: mmm packages/modules/Bluetooth

If during stream configuration, remote device rejects any of the
configuration stream, instead of disconnecting link as a result of not
moving forward in the ASE state machine, Android will just drop creating
stream and notify audio framework about that

Bug: 294943711
Test: atest bluetooth_le_audio_test
Tag: #feature
Change-Id: I10b13448a8d43da8f13ced6ffa0b79dfc704ddc6

If remote device indicates a Context Type as supported but it is not in
Available context types, such stream will not be started by the Android

Bug: 281957456
Tag: #feature
Test: atest bluetooth_le_audio_client_test

Change-Id: I71f099aff1f93208d405143cfb8134b8be4bda79

With this patch available group contexts and locations are updated in following
points of time
1) When device got disconnected
2) When notification arrives but group is not streaming
3) when streaming ends
4) when device got connected

Also note, that whenever leaudio code get latest available context type
it just look into connected devices

Bug: 281957456
Tag: #feature
Test: atest bluetooth_le_audio_client_test

Change-Id: I7a86c6520a708291f2f05c1750c005732bc379d6

Cleanup in the context of the context type work

Bug: 281957456
Tag: #feature
Test: atest bluetooth_le_audio_client_test

Change-Id: Ic91d2cf515ff0a47ecde70de1879147caae99850

This is needed for future tests

Bug: 281957456
Tag: #feature
Test: atest bluetooth_le_audio_client_test

Change-Id: I94b49fe2b95653e261ef02839342764599852657

Bug: 285791484
Test: N/A
Tag: #refractor
Change-Id: I0e19576644e9b471fdcf31f0b2ff9c7c0ee4c8dd

When the Bluetooth Quality Report Event came too short the process
crashed. As per current implementation there is a possibility
that a BQR event is shorter, because not all controllers' vendors
extended the BQR event. A temporary solution is to only print
the error and return.

Bug: 293849013
Test: m libbtif-core
Test: manual - check BQR event logs
Change-Id: Iab7ec1f83135ddf8d517bbe919d3c9df1648885a

This change introduces support for fallback to Unicast streaming once
Broadcast streaming is finished. To make this happend Broadcast needs to
expropriate Unicast stream then this Broadcast needs to be stopped.

Tag: #feature
Bug: 294513510
Test: atest LeAudioServiceTest
Change-Id: I2cdd8f2576b2a1bdd30e1c8fd8cff870cd4be552

In handover scenario from Unicast to Broadcast there is need to wait
until Unicast would be INACTIVE before setting Broadcast.

Tag: #feature
Bug: 286598970
Test: atest LeAudioServiceTest
Change-Id: I8baede89a34bacb6e36a0f182522c56276cb48b9

This functionallity is provided by ASAN instead.
This change also enable ASAN in tests that where
using the AllocationTestHarness.

Test: mmm packages/modules/Bluetooth
Bug: 295430884
Change-Id: Ied955493ea456bc8920fb8661f444c62a6e07a3c