Bug: 231345733
Test: mma -j $(nproc)
Test: ./build.py
Tag: #floss
Change-Id: I649216d42d4bb204845941c5fba738ff27e43a3a

Bug: 231345733
Test: mma -j $(nproc)
Test: ./build.py
Tag: #floss
Change-Id: Idcfa21b8aeffa63ac3a886d0ba8b489802fc3485

Bug: 231345733
Test: mma -j $(nproc)
Test: ./build.py
Tag: #floss
Change-Id: Ibe5b65072a349b2fe5cc54c6d66e6e3abc2871eb

Needed to plumb the suspend_id from powerd all the way to the suspend
function in the suspend.rs module.  Then we have access to the suspend
id.

Also, convert the types from u32 to i32 to match the protobuf

Bug: 231570916
Test: ./build.py
Test: mma -j $(nproc)
Test: system/gd/cert/run --topshim
Tag: #floss
Change-Id: Id71ee50035441320e477b2d8f805a73e5edd696a

Bug: 231435700
Test: ./build.py
Test: mma -j $(nproc)
Test: system/gd/cert/run --clean --topshim SuspendTest
Tag: #floss
Change-Id: Ia4324c4e95ee5fd5313029fdd7cecd1142c45b0c

Bug: 232547719
Test: mma -j $(nproc)
Test: ./build.py
Test: system/gd/cert/run --clean --topshim
Tag: #floss
Change-Id: I996cf9e49f8324f07f3c6d7c6fd78369532a8f16

* changes:
  LeAudioTestApp: Add button to trigget set active group
  HapClientService: Don't use cache for PTS
  has_client: Add upper tester for the PTS
  has_client: Disconnect profile when control point is not accesable
  has_client: Improve subscribing for notification
  has_client: Fix Indication confirmation
  has_client: Always write CCC values
  has_client: Improve subscribtion to control point
  gatt: Fix incorrect GATT confirmation on indication

- Adds OnPairing method to receive a stream of pairing events.
- Reorganizes interface.

Test: m PandoraServer
Change-Id: I578542dca3f5e9cced0c6a53f1f693c62119a4da

This is needed for PTS

Bug: 238171211
Test: tested manually with PTS
Tag: #feature
Change-Id: I72de83488e8714359be113c25f459e8de2dcda37

PTS wants us to do operations on remote device, so when enabled, call
native.

Bug: 238171211
Test: atest BluetoothInstrumentationTests
Test: PTS
Tag: #feature
Change-Id: I14f8eda50321f941b69fa1cd37403b131dc5938d

With this patch, if cache is off, we read all presets to satisfy PTS.

Bug: 238171211
Test: manualy with PTS
Test: atest BluetoothInstrumentationTests
Tag: #feature
Change-Id: I641e9933d0a967c4949ff3374e8f43d119fe3b92

This is done for PTS which is not ready to accept control point
commands and some time.
But in the real live, we actually disconnect such device if there is
problem with control point operations. We can always reconnect if
needed.

Bug: 238171211
Test: atest BluetoothInstrumentationTests
Test: PTS
Tag: #feature
Change-Id: Iec5411267ef9c03368a99b89ab6e9fab8a96200e

To improve PTS testing, let's subscribe to all the CCC before making
operations on the control point.

Bug: 238171211
Test: manualy with PTS
Test: atest BluetoothInstrumentationTests
Tag: #feature
Change-Id: Iebb66edff13edea336de32292fc54f311f72d4ed

Bug: 238171211
Test: PTS
Test: atest BluetoothInstrumentationTests
Tag: #feature
Change-Id: Iba9fcfe6926a137f787f7fa4aa230153dde9998f

We do not trust devices and always write CCC.
It will also help with PTS

Bug: 238171211
Test: atest BluetoothInstrumentationTests
Test: PTS
Tag: #feature
Change-Id: Ifb354d772fcdf6963f24e962e8b2864902bfb8f2

Tag: #feature
Test: atest pts-bot:GATT/CL/GAD
Bug: 239103399
Change-Id: Ie8533203eb061177792517935a7963aa4171f99f

This change adds functionality to detect whether the
host machine is x86_64, which is the only machine currently supported by
the build script.

Bug: 241967798
Tag: #floss
Test: run ./build.py on non x86_64 and observe script exiting, run on x86_64 and
observe normal behavior.

Change-Id: Ia9837e26c78559e85b00cc8637daa2874cb3d8d8

* changes:
  LeAudioTestApp/Bass: Allow channel selection
  Bass: Fix BIS sync state channel map

Test: atest avatar
Change-Id: I59c1a6f6ff87525fde7c7d0e102d89e739fc4247

Bug: 242054540
Test: gd/cert/run
Tag: #refactor
BYPASS_LONG_LINES_REASON: Bluetooth likes 120 lines

Change-Id: I1412d795330c5ee593daab82efdc3d96324f6f97

Bug: 241871296

The bug is that mSignalStrengthUpdateRequest is null when bt calls mTelephonyManager.clearSignalStrengthUpdateRequest(mSignalStrengthUpdateRequest).

mSignalStrengthUpdateRequest is created at constructor. So the only explanation is that there is a race condition between the constructor and stopListenForPhoneState() who calls clearSignalStrengthUpdateRequest();

Test: atest BluetoothInstrumentationTests
Change-Id: I0d456a4a7981e1befe291d2045b926a4da4b6bf5

Rename the python parser to scripts/pdl/parser.py,
and the test file to tests/python_parser_test.py

Test: atest --host pdl_python_parser_test
Change-Id: Ie32e9e384810bbc991e291e4863d37845f73c9b7

Some types had explicitly defaulted constructors even though they
are not actually default-constructible. Also fixes an unused
variable warning.

Tag: #refactor
Bug: 241680050
Test: atest --host bluetoothtbd_test net_test_osi
Change-Id: Ieea664fe7c6ecdca6be352777c775dfd4649e5bd

packages/modules/Bluetooth/android/app/Android.bp include
PlatformProperties as dependency as well as
"framework-bluetooth-pre-jarjar". The last one also adds the
dependency PlatformProperties, which it is causing the test
StrictJavaPackagesTest.testApkInApex_nonClasspathClasses to
fail.

Note:

The problem the test is trying to detect is multiple copies
of the same class across the bootclasspath; in this case,
framework-bluetooth is part of the bootclasspath, so all its
classes are by default available (with caveats) to all java code;
PlatformProperties is being bundle again in the Bluetooth APK
(and applying the same jarjar rules as framework-bluetooth),
so the end result is that we are bundling the same class that
is already in the bootclasspath.

Bug: 241475987
Tag: #compatibility
Test: Run Bluetooth test cases.

Change-Id: Ia7015fe1dcd279b5e39853a234fef7d908eb8236
Merged-In: Ia7015fe1dcd279b5e39853a234fef7d908eb8236

500 ms should be sufficient for finding the Bluetooth HAL,
but when it is not, complain loudly and continue, in case
the HAL still starts.

Bug: 233804549
Test: atest
Change-Id: Ic20f0bc4f494f8e72aec1af8205b7be5e0ce9165

As per specification we should register for notification when EATT is
used and indication if EATT is NOT used. This is bit odd as having
support and being able to connect, and maintaing connection are
different things.
To be on the save side we would like to register to both - notification
and indication and let server to decide what it is sent (based on the
bearer).

With this patch, we register to anything what servers allow dispite
EATT tracking.

This will also help PTS testing

Bug: 238171211
Test: atest BluetoothInstrumentationTests
Test: PTS
Tag: #feature
Change-Id: I7c30278de00e3e75c6ecd9b6c55258b7eef92fe9

If there is many GATT users, Android sends indication to all the users
and expects confimation.
Without this fix, we could endup sending multiple Confirmation to remote
device on single Indication - which is wrong.

This patch fixes this.
Found during PTS HAP testing.

Bug: 238171211
Test: atest BluetoothInstrumentationTests
Test: PTS
tag: #feature
Change-Id: Ieef9a0f3448ce8529624e777282e857a7f206c00