Hi Luiz, >Hi Claudia, > >On Wed, Jul 12, 2023 at 10:45 AM Luiz Augusto von Dentz <luiz.dentz@xxxxxxxxx> wrote: >> >> Hi Claudia, >> >> On Wed, Jul 12, 2023 at 5:46 AM Claudia Draghicescu >> <claudia.rosu@xxxxxxx> wrote: >> > >> > This series of patches adds support for BAP broadcast sink. >> > It consists in registering a sink endpoint using the Sink PAC UUID, >> > discovering of broadcast advertisers that announce the Broadcast >> > Audio Announcement service, synchronizes to the Periodic >> > advertisements of the source and synchronizes to the BIG advertised >> > in the PA train. >> > To retrieve the BASE info advertised in the PA train, the patch >> > Bluetooth: ISO: Add support for periodic adv reports processing was >> > used. >> > >> > This feature was tested using bluetoothctl with the following commands: >> > >> > [bluetooth]# endpoint.register 00002bc9-0000-1000-8000-00805f9b34fb >> > 0x06 [/local/endpoint/ep0] Auto Accept (yes/no): y >> > [/local/endpoint/ep0] Max Transports (auto/value): a >> > [/local/endpoint/ep0] unicast/broadcast (u/b): b >> > [/local/endpoint/ep0] BIG (auto/value): a [/local/endpoint/ep0] BIS >> > (auto/value): a >> >> Ive thought we would be using BAA UUID instead to follow the same >> setup as the broadcaster role. We think adding a new UUID for the broadcast sink (0x1851) along with the UUID for the broadcast source (0x1852) is the best choice because we need to differentiate the endpoint and transport in other applications like PipeWire. This follows the same procedure as for bap unicast where PAC_SOURCE_UUID and PAC_SINK_UUID are used to differentiate between BAP unicast source and sink endpoints and transports. >> >> > [bluetooth]# scan on >> >> Then while scanning if we find an announcement it should create a >> remote endpoint under the device object path, just as in case of >> unicast, which can later be used with endpoint.config. > >I had a few more thoughts about how to enumerate the remote broadcast endpoints, we may want to introduce a driver for BAA UUID so when a device is found with BAA as part of the advertised >UUID would get probed and the driver can take care of creating the MediaEndpoint objects based of the advertised data, probably each BIS should have a endpoint, > >> > [bluetooth]# endpoint.config /org/bluez/hci0/pac_bcast0 >> > /local/endpoint/ep0 16_2_1 <source_address> >> >> Here then one would replace /org/bluez/hci0/pac_bcast0 with the one >> found during the scanning, so the difference with broadcaster becomes >> the source endpoint, if you pass the adapter it is for broadcaster >> role or in case it is a remote endpoint the it is for listener role so >> the commands would work similarly in all directions >> >> > Claudia Draghicescu (6): >> > client/player: Add broadcast sink endpoint registration and >> > configuration. >> > client/main: Add broadcast source discovery >> > media: Add support for a broadcast sink media endpoint >> > transport: Update transport properties for a broadcast stream >> > btio: Add support for getsockopt(BT_ISO_BASE) >> > bap: Add gdbus interface for BAP broadcast sink, create >> > synchronization with source and create BAP broadcast sink stream >> > >> > btio/btio.c | 13 +- >> > client/main.c | 57 +++++- >> > client/player.c | 187 +++++++++++++++++++- >> > client/player.h | 3 + >> > profiles/audio/bap.c | 347 ++++++++++++++++++++++++++++++++++--- >> > profiles/audio/media.c | 35 +++- >> > profiles/audio/media.h | 2 +- >> > profiles/audio/transport.c | 244 +++++++++++++++++++++++++- >> > src/shared/bap.c | 50 +++++- >> > src/shared/bap.h | 4 +- >> > 10 files changed, 888 insertions(+), 54 deletions(-) >> > >> > >> > base-commit: 838e1578072900d1f98dfb31cc538940d2fad876 >> > -- >> > 2.34.1 >> > >> >> >> -- >> Luiz Augusto von Dentz > > > >-- >Luiz Augusto von Dentz
