* Never set promise more than once
* To verify running task cannot be cancelled, make sure Cancel() is
  called after callback starts
* Remove a stress test case. Sometimes there is unexpected event causing
  deviation.

Test: run Timer test for 5000 times
Bug: 116081383
Change-Id: If99e4958289e9b6aaabc51b15ce4ac035a3ff1e9

Test: A2DP stream music
Change-Id: I4f7325b3eb3be07eca6c599c7f6f5e037b8e7a64

When sending Le Start Encryption, it's possible that at the same exact
moment the connection was disconnected, and there is an incoming
"Disconnection Complete Event". The controller would send error 0x02
"unknown connection identifier" in such case. This can happen
sporadically when reconnecting device multiple times.

Currently, we don't handle this error code. This lead to the timeout
when waiting for "Encryption Complete", and unbonding of the device.

From now on, if 0x02 error code is returned, we'll cancel current SMP
operation, which is the "LE Start Encryption". This will stop the SMP
timer, and make sure the bond to remote device is not lost.

Test: Manual test reconnecting with HID device multiple time in a row,
causing the link drop right after connection, by removing battery, or
closing the shield box.
Bug: 113652889
Change-Id: I2ff9c13dbc8e7b71505908996e26b89fa1ea6a42

In Bluetooth native infrastructure refactor plan, we need a state
machine class for entire stack

Test: manual
Change-Id: I0e6671e42570482d9a9466b117054b3aadb3e9cc

When shutting down a2dp source,
A2DP source set StateShuttingDown state too fast cause the A2DP doesn’t stop stream.
Will try to stop A2DP when doing A2DP source end session.

Bug: 112279820
Test: Build Pass, BT off/on when play A2DP
Change-Id: I6bfb5c235da1ff4a7e6814bae3a98ae7dd1788aa

During StartUp/ShutDown(), if another running task in message loop needs
api_lock, then it will deadlock. Release api_lock in StartUp and
ShutDown when current thread is blocked by another task

Bug: 116081383
Test: run unit test
Change-Id: Ic34406d0cfc6364ac965c4f1aa917a5ccaf6b373

Also increase tolerance for unit tests

Bug: 116081383
Test: Run unit test
Change-Id: I5531c2259cbbc58b6ee7527813e8d5c4164e94c5

Adjusted the requested data length to the BT Controller from 168 to 167.
Also, change the default media type during playback to unknown.

Test: Manual testing with Hearing Aids and verifying with OTA logs
Change-Id: Ic4e934729a98bef5c10ebca64093d8a192e79a01

They are not related to GATT, update naming.

Test: compilation
Bug: 67058417
Change-Id: Ia84ed98e02fbc04ecfde979f860b8766b0018218

When user requests Disconnect(), after disconnecting GATT,
OnGattDisconnected() never cleans up GATT queue, because the
corresponding device is already removed from HearingDevices. We need to
clean up the queue before connecting the same device again. Without the
fix, sometimes ACL is connected, but the host failed to read GATT
characteristic to open COC. Also clean up connection update status
during disconnect

Bug: 115365334
Test: connect to both devices -> disconnect -> connect, and repeat
Change-Id: I158b19b3d1f2cb2123ec554eb1ffe8c842df998e

* Add "postsubmit" test group to TEST_MAPPING in system/bt
* Developer can run tests mannually in system/bt via:
  system/bt$ atest :postsubmit
* Remove legacy configuration files spread across system/bt
* Created unified configuration file system/bt/AndroidTestTemplate.xml
  and add it to the default build config so that developers don't have
  to configure it

Bug: 113359343
Test: atest :postsubmit
Change-Id: I0fdd55b47ef176b5c4ce3d2cd217ffa67cf9bc23

For host conn_id, use hex format

Test: manual
Change-Id: I7d7e690a1dbe0665b827c59d104540eafa45d426

Timer conn.process_repage_timer is being freed in HID_HostDeregister
which causes use after free condition in HID_HostRemoveDev->
HID_HostCloseDev.

Fix:
Call alarm_free after HID_HostRemoveDev in HID_HostDeregister

Test:
Code analysis

Bug: 116774368
Change-Id: Iddecd126716a5660f4c37ddc917bacee84342297

C++ 17 has a bunch of features that will help us refactor the Bluetooth
stack to be much more safe and stable.

Bug: 115895445
Test: Compile and testtracker/201115
Change-Id: Ia1e8e179fa1df116969606d119cb8abe11f44761

* Add a new SSR entry BTA_DM_PM_SSR4 with maximum
  remote/local timeout value.
* Use BTA_DM_PM_SSR4 when A2DP is streaming to prevent
  phone enter sniff subrating mode.

Bug: 113496412
Test: Manually
Change-Id: I194ef9fdb4dd38eab67ff20e1bb9cab47edbfe8f

That field is documented in Bluetooth MAP 1.4 spec, page 114.
Spec is here: http://shortn/_zfoIrfbIzR

Bug: 111614861
Test: tested w/ KitchenSink App
Change-Id: I89903c386d97934ec7b9c80333ecc6a6d560cbec

Bug: 116614593
Test: build with WITH_TIDY=1
Change-Id: I4adaf89db5001e9b627beac8160f62b50513d18f

Bug: 27078725
Test: Code compilation
Change-Id: I9398f0354e54cc6a754ed46b05b9f7cef09add8f

* changes:
  Timer: Don't run new task at old timepoint
  Timer: Separate code path for single and periodic

* Fix a race condition that the task is just finished but IsScheduled
  is true
* Fix a bug where cancelling the old task will make new task run at
  the previous task's scheduled timepoint
* Protect racing read/write to message_loop_thread_ in RunTask() and
  Cancel() by storing a local reference to message_loop_thread_ in
  RunTask()

Bug: 116081383
Test: Run bluetooth_test_common.TimerTest
      reschedule_task_dont_invoke_new_task_early
      schedule_task
      for multiple times
Change-Id: I737a864e1b061adbcd41245711e0d3e8adf448d4

The volume is in the Start command already, no need to send it in a
separate command.

Bug: 116145308
Test: play music, pause, adjust volume, start playing again, verify
      playing was re-started with proper volume level.
Change-Id: I2a900b03a35670de4a265184f333517db17497d8

* Add helper RunSingleTask() and RunPeriodicTask() to handle two cases
  separately
* In RunTask(), check that it must run on specified message loop thread

Bug: 116081383
Test: Run bluetooth_test_common
Change-Id: Idc41b2a509f43dbe946b2e26f2afd775726514aa

* Use more efficient overloaded string methods.

Bug: 30411878
Test: build with WITH_TIDY=1
Change-Id: Iff759bf68ede3bb6bdd8e332b3490128eee3b5fe

Add FALLTHROUGH_INTENDED for clang compiler.

Bug: 112564944
Test: build with global -Wimplicit-fallthrough.
Change-Id: Iedf7e46445101a5e3358473e1200212d67cf58d1

This is the first step for it to run in lab. With this, the test binary
will be included in general-tests.zip (test_suite* target). From there,
we can kick off the test in the lab using TF, with the right options.

Bug: 111921389
Test: m -j net_test_avrcp
./host/linux-x86/testcases/net_test_avrcp/x86_64/net_test_avrcp

Change-Id: I4de9409b4a9d4c066a9024a33d6a980ecb451dc1