Bug: 237447510
Test: atest pts-bot:HFP
Change-Id: Ie16aa3ed8fa6c49afec1e8814dd228320fb9be8e

- Add definition of SetEventMaskPage2 in hci_packets.pdl
- Use generated class for SetResolvablePrivateAddressTimeout
- Remove unecessary validation of LeIsoCommand views

Test: m root-canal
Change-Id: I223c9f56839a734f6b11e37a7ae3417ef9c25153

- Remove some declared but un-implemented functions
  in DualModeController
- Move suggested_max_tx_octets and suggested_max_tx_time
  to the link layer controller
- Change the type of extended inquiry response
  to std::array<uint8_t, 240>, which accurately
  reflects the length of the response as passed
  from the HCI command

Test: m root-canal
Change-Id: I8837333f9c080454d987a9b7877d8c55e6d0c324

This pins all the lmp dependencies to a set of crates that we know that
we have.

Bug: 238726856
Test: Build succeeds
Change-Id: I35cf5198fe7af22ceac2dde3f9023608c6f0a7c7

This macro is provided by the gd header os/include/osi.h.
This change is made to remove dependencies from Rootcanal on gd's
runtime.

Bug: 256013143
Test: m root-canal
Change-Id: I5a003a6885cad58c7421c58527524c8a3ea36d3b

The initialisers in controller_properties.cc make a good
default configuration.

Test: m
Change-Id: I3b90acb61c11d1518ca8be4a600f97d5563d28a1

Bug: 244451007
Test: atest rootcanal_hci_test
Change-Id: Id3e34b940ed87c226866294a1519d2e575ccc5e9

Clap has released a version 4, but Android is currently on version 3.
The update in https://r.android.com/2241208 breaks if Clap 4 is used
by accident. By adding the current version numbers, we can guard
against these problems.

Test: cargo update && cargo test
Change-Id: I270adb4538a83748d477ee6091bc4a2605183f48

The section Vol 2, Part C § 3.5 Feature requirements defines
requirements for the LMP feature mask. This change checks all
of these and emits warning for invalid features.

Test: m root-canal
Bug: 253524415
Change-Id: I13db270f7748003f4f0e1e22cd6f335dabcb97a6

Specifically the checks run by the initiator before sending a
connect indication to a connectable advertising PDU.

Test: gdcert + pts-bot
Change-Id: I759cbde63505528b9ba296835a542b40b0c425c4

- Implement HCI parameter validation for LE Create Connection and
  LE Extended Create Connection commands
- Group the initiator state parameters into a sub-class
  LinkLayerController::Initiator

The LE Connection logic in ConnectIncomingLegacyAdvertisingPdu and
ConnectIncomingExtendedAdvertisingPdu is unchanged

Test: atest rootcanal_hci_test
Change-Id: I5db4372effbb373aa994a6dca91e212e373fd151

The parameter can be used to configure the prefix for all
Bluetooth addresses generated from a RootCanal instance, so as
to disambiguate addresses generated from multiple connected
instances.

Bug: 254735655
Test: m root-canal + presubmit
Change-Id: Idda6c6dc56f5e56099df9079724587b3bef2193e

Test: logging only
Change-Id: I6b476cf0122a8b2ac9280384d217d6b6754b488e

This reverts commit 756c0bd9.

Reason for revert: Introduce PTS-bot test failures

Change-Id: Icca6717578dcdbb27bd43eb9aa9eca44bcae51eb

Configure the capabilities with version v1.0, and all
additional features disabled.

Test: None
Change-Id: I8e8823e345ecd2494ec39b71dce890a0e22daaba

Refactor LE extended advertising to implement HCI parameter
validation as required by the specification.

Add a new link layer packet type LE_EXTENDED_ADVERTISING_PDU
corresponding to ADV_EXT_IND, AUX_ADV_IND, AUX_CHAIN_IND PDUs,
and implement proper handling for extended scanning and legacy
initiator states

Test: atest --host rootcanal_hci_test
Test: gd/cert
Test: pts-bot
Change-Id: I6696ccc47de7b72715e6f6eeda3ab4c5511e11af

The new python path behavior has been enabled by default,
so enabling it manually with modern_python_path_defaults
is no longer necessary.

Bug: 245583294
Test: Presubmits
Change-Id: I0d2a15216be9481f3851a054351c20fbbbde8b5f

47229aab

Test: atest pts-bot
Change-Id: Iad30acd30bd0a6256514bf6ed17d96f03fe144cd

98ab2bf0

Test: atest pts-bot
Change-Id: I290b98f6d39da772e18ad4d29a7fe646e4d6effc

Revert submission 2234606

Reason for revert: Breaks pts-bot presubmit, resolution takes longer than expected
Reverted Changes:
I83d2bbad3:RootCanal: Refactor legacy LE advertising commands...
Ice2e2e46b:RootCanal: Refactor LE scanning commands

Test: pts-bot presubmits
Change-Id: I3cdee05ee868d38bae411767d898f3f213cf93e0

Revert submission 2234606

Reason for revert: Breaks pts-bot presubmit, resolution takes longer than expected
Reverted Changes:
I83d2bbad3:RootCanal: Refactor legacy LE advertising commands...
Ice2e2e46b:RootCanal: Refactor LE scanning commands

Test: pts-bot presubmits
Change-Id: I5378389a4713627d1216e7f8f36b75870669c68a

- Add tests with min and max scalar values
- Add tests for array with a size field and
  dynamically sized elements
  (real life example TLV buffers)
- Fix python generation bug with said arrays

Test: atest --host pdl_python_generator_test
Change-Id: Iad255cc452399967e24e212a82aff747fca3b37e

* DirectAdvertisingAddressType::NO_ADDRESS{=>_PROVIDED}
* PacketStatusFlag::NO_DATA{=>_RECEIVED}

libchrome started to include libevent.h from r1033946
(crrev.com/c/3811220) and hence indirectly include netdb.h, which
defines macro NO_DATA and NO_ADDRESS. This causes compile error in floss
when they are used in the enums.

Use NO_ADDRESS_PROVIDED and NO_DATA_RECEIVED to be consistent with the
spec (downloadable from
https://www.bluetooth.com/specifications/specs/core-specification-5-3/,
page 2272 and 1814 respectively)

BUG: 253138188
Tag: #floss
Test: emerge-hatch libchrome floss
Change-Id: I493017f800ffddab2a24d8cf3324d8e21e96aee0

Test: atest pdl_tests pdl_inline_tests
Change-Id: I6cd6ff18b75ab3b40c21ac87cae14a04e816beef

Test: atest pdl_tests pdl_inline_tests
Change-Id: I686e0558b10d7e0c644d70f0cb1061e86a33a624

Test: atest pdl_tests pdl_inline_tests
Change-Id: I2532bf2b90b8bd8107070cac5782b6d15c15ee54

Test: atest pdl_tests pdl_inline_tests
Change-Id: I11416277d43ef8ebb78608e7403f5e63636dd02c

We should actually depend on the same libraries as for the rest of the
binary. This allows us to more freely move test code around.

Test: m pdl_tests
Change-Id: I97bdd2bf469ada187357ee58cd42b9095c6f0681

Test: atest pdl_tests pdl_inline_tests
Change-Id: I46bf0c83bfaf51165cd2ee6c699bc6032b38de44

This requires more spurious ‘ast::Field’ -> ‘Field’ conversions, which
we’ll clean up next.

Test: atest pdl_tests pdl_inline_tests
Change-Id: I5f639a94e164696c311edbb54b77c10135f3a043

Test: atest pdl_tests pdl_inline_tests
Change-Id: Ib4ff14eb44b9b4384e6c69c2d15410060eaa6925

The ‘Field::from’ calls will eventually go away when we moved more
code to ‘Field’.

Test: atest pdl_tests pdl_inline_tests
Change-Id: I59c1d4c42fa47e512e037d7219ffa937a15c2be3

I’ll slowly be moving over functionality to this new enum and its
variants.

While this might seem like a lot of boiler-plate, this simplifies the
code overall since structs like ‘ScalarField’ know exactly what type
they are. So we’ll have fewer redundant match expressions after this.

Test: atest pdl_tests pdl_inline_tests
Change-Id: Ib9147d4899944e74d2e06bd41e3adc832f89e802

Implement HCI parameter validation as required by the specification.
Resolve or generate private addresses when required in
6.B.6.3 Privacy in scanning state.

Test: atest --host --no-bazel rootcanal_hci_test
Change-Id: Ice2e2e46b8ca157ca1b4131d0c7a642a2074ddc6

Migrate the pdl tool from structopt to clap 3 and use the derive API.

Bug: 241573790
Test: m pdl + run `pdl --help`
Change-Id: I23fa9220208afa9e0832f94b1a25d3f13dd40d00

The python path for soong-built modules used to automatically
include all the subdirectories of the root package, and also
the package of the entrypoint script. Soong is being updated
to remove that, to match bazel's behavior and soong's behavior
with embedded_launcher: true, so update these scripts to work
with the new behavior.

Bug: 245583294
Test: m pdl_python_generator_test pdl_python_parser_test && /ssd/aosp-master/out/host/linux-x86/nativetest64/pdl_python_generator_test/pdl_python_generator_test && /ssd/aosp-master/out/host/linux-x86/nativetest64/pdl_python_parser_test/pdl_python_parser_test
Change-Id: I6b64c0ed77e41348992cf266ce601977ac754f20

On an M1 the writer thread can write faster than the reader can read.
This results in the reader never exiting the loop, causing the test to
hang.

We now lower the rate of writing to make sure the readers can always
keep up.

Bug: 251187705
Test: Unit tests now succeed on M1
Change-Id: Ia675c3b3b0613b2b00a588a803473a3d997bf0ee

There are multiple duplicated compiler options
configs for turning warnings into error.
This patch factor out these duplicated configs

Bug: 250720917
Test: refactoring CL. Existing unit tests still pass.

Change-Id: I66ace35eea122be2458b6eb03ac406c8ba50ca61

Implement HCI parameter validation as required by the
specification. Resolve or generate private addresses
when required in 6.B.6.2 Privacy in advertising mode.

Test: gd cert
Test: atest --host --no-bazel rootcanal_hci_test
Change-Id: I83d2bbad31ec41c4c8b5c27c6280530bfb8dbb6a

Fix: 247975614
Test: atest --host lilmp_tests
Change-Id: I70d6272c8ab34a8258655ea9d11b597dc82865a4