Hi Pauli,
On Tue, Feb 25, 2025 at 5:10 PM Luiz Augusto von Dentz
<luiz.dentz@xxxxxxxxx> wrote:
>
> Hi Pauli, Bastien,
>
> On Tue, Feb 25, 2025 at 5:01 PM Luiz Augusto von Dentz
> <luiz.dentz@xxxxxxxxx> wrote:
> >
> > From: Luiz Augusto von Dentz <luiz.von.dentz@xxxxxxxxx>
> >
> > This adds support for PreferredBearer which is printed with the likes of
> > info command:
> >
> > bluetoothctl> info <addr>
> > ...
> > PreferredBearer: last-seen
> >
> > It also introduces a new command to get/set the PreferredBearer:
> >
> > [bluetoothctl]> bearer --help
> > Get/Set preferred bearer
> > Usage:
> > bearer <dev> [last-seen/bredr/le]
> >
> > [bluetoothctl]> bearer <addr>
> > PreferredBearer: last-seen
> > [bluetoothctl]> bearer <addr> le
> > bluetoothd: @ MGMT Command: Add Device (0x0033) plen 8
> > LE Address: <addr>
> > Action: Auto-connect remote device (0x02)
> > [CHG] Device <addr> PreferredBearer: le
> > Changing le succeeded
> > [bluetoothctl]> bearer <addr>
> > PreferredBearer: le
> > [bluetoothctl]> bearer <addr> bredr
> > bluetoothd: @ MGMT Command: Remove Device (0x0034) plen 7
> > LE Address: <addr>
> > [CHG] Device <addr> PreferredBearer: bredr
> > Changing bredr succeeded
> > ---
> > client/main.c | 27 +++++++++++++++++++++++++++
> > 1 file changed, 27 insertions(+)
> >
> > diff --git a/client/main.c b/client/main.c
> > index feb21a1163d2..76c9bc329c96 100644
> > --- a/client/main.c
> > +++ b/client/main.c
> > @@ -1714,6 +1714,7 @@ static void cmd_info(int argc, char *argv[])
> > print_property(proxy, "AdvertisingFlags");
> > print_property(proxy, "AdvertisingData");
> > print_property(proxy, "Sets");
> > + print_property(proxy, "PreferredBearer");
> >
> > battery_proxy = find_proxies_by_path(battery_proxies,
> > g_dbus_proxy_get_path(proxy));
> > @@ -2086,6 +2087,30 @@ static void cmd_wake(int argc, char *argv[])
> > return bt_shell_noninteractive_quit(EXIT_FAILURE);
> > }
> >
> > +static void cmd_bearer(int argc, char *argv[])
> > +{
> > + GDBusProxy *proxy;
> > + char *str;
> > +
> > + proxy = find_device(argc, argv);
> > + if (!proxy)
> > + return bt_shell_noninteractive_quit(EXIT_FAILURE);
> > +
> > + if (argc <= 2) {
> > + print_property(proxy, "PreferredBearer");
> > + return;
> > + }
> > +
> > + str = strdup(argv[2]);
> > +
> > + if (g_dbus_proxy_set_property_basic(proxy, "PreferredBearer",
> > + DBUS_TYPE_STRING, &str,
> > + generic_callback, str, free))
> > + return;
> > +
> > + return bt_shell_noninteractive_quit(EXIT_FAILURE);
> > +}
> > +
> > static void cmd_list_attributes(int argc, char *argv[])
> > {
> > GDBusProxy *proxy;
> > @@ -3247,6 +3272,8 @@ static const struct bt_shell_menu main_menu = {
> > dev_generator },
> > { "wake", "[dev] [on/off]", cmd_wake, "Get/Set wake support",
> > dev_generator },
> > + { "bearer", "<dev> [last-seen/bredr/le]", cmd_bearer,
> > + "Get/Set preferred bearer", dev_generator },
> > { } },
> > };
> >
> > --
> > 2.48.1
>
> So I went ahead and implemented the idea of having PreferredBearer,
> this works great when setting bredr it really stops from connecting to
> LE, the said the other way around when setting to le seems to confuse
> some headsets like EarFun and it ends up connecting both bearers:
>
> [EarFun Air Pro 3]> transport.show
> Transport /org/bluez/hci0/dev_70_5A_6F_63_B6_41/pac_source0/fd1
> UUID: Sink PAC (00002bc9-0000-1000-8000-00805f9b34fb)
> Codec: 0x06 (6)
> Configuration.#0: len 0x02 type 0x01
> Configuration.Sampling Frequency: 16 Khz (0x03)
> Configuration.#1: len 0x02 type 0x02
> Configuration.Frame Duration: 7.5 ms (0x00)
> Configuration.#2: len 0x05 type 0x03
> Configuration.Location: 0x00000001
> Configuration.Location: Front Left (0x00000001)
> Configuration.#3: len 0x03 type 0x04
> Configuration.Frame Length: 30 (0x001e)
> Configuration.#4: len 0x02 type 0x05
> Configuration.Frame Blocks per SDU: 1 (0x01)
> Device: /org/bluez/hci0/dev_70_5A_6F_63_B6_41
> State: idle
> Volume: 0x00c8 (200)
> Endpoint: /org/bluez/hci0/dev_70_5A_6F_63_B6_41/pac_source0
> QoS.CIG: 0x00 (0)
> QoS.CIS: 0x00 (0)
> QoS.Framing: 0x00 (0)
> QoS.PresentationDelay: 0x00009c40 (40000)
> QoS.Interval: 0x00001d4c (7500)
> QoS.Latency: 0x0008 (8)
> QoS.SDU: 0x001e (30)
> QoS.PHY: 0x02 (2)
> QoS.Retransmissions: 0x02 (2)
> Location: 0x00000003 (3)
> Links: /org/bluez/hci0/dev_70_5A_6F_63_B6_41/pac_sink0/fd3
> Transport /org/bluez/hci0/dev_70_5A_6F_63_B6_41/pac_source0/fd2
> UUID: Sink PAC (00002bc9-0000-1000-8000-00805f9b34fb)
> Codec: 0x06 (6)
> Configuration.#0: len 0x02 type 0x01
> Configuration.Sampling Frequency: 16 Khz (0x03)
> Configuration.#1: len 0x02 type 0x02
> Configuration.Frame Duration: 7.5 ms (0x00)
> Configuration.#2: len 0x05 type 0x03
> Configuration.Location: 0x00000002
> Configuration.Location: Front Right (0x00000002)
> Configuration.#3: len 0x03 type 0x04
> Configuration.Frame Length: 30 (0x001e)
> Configuration.#4: len 0x02 type 0x05
> Configuration.Frame Blocks per SDU: 1 (0x01)
> Device: /org/bluez/hci0/dev_70_5A_6F_63_B6_41
> State: idle
> Volume: 0x00c8 (200)
> Endpoint: /org/bluez/hci0/dev_70_5A_6F_63_B6_41/pac_source0
> QoS.CIG: 0x00 (0)
> QoS.CIS: 0x01 (1)
> QoS.Framing: 0x00 (0)
> QoS.PresentationDelay: 0x00009c40 (40000)
> QoS.Interval: 0x00001d4c (7500)
> QoS.Latency: 0x0008 (8)
> QoS.SDU: 0x001e (30)
> QoS.PHY: 0x02 (2)
> QoS.Retransmissions: 0x02 (2)
> Location: 0x00000003 (3)
> Links: /org/bluez/hci0/dev_70_5A_6F_63_B6_41/pac_sink0/fd4
> Transport /org/bluez/hci0/dev_70_5A_6F_63_B6_41/pac_sink0/fd3
> UUID: Source PAC (00002bcb-0000-1000-8000-00805f9b34fb)
> Codec: 0x06 (6)
> Configuration.#0: len 0x02 type 0x01
> Configuration.Sampling Frequency: 48 Khz (0x08)
> Configuration.#1: len 0x02 type 0x02
> Configuration.Frame Duration: 7.5 ms (0x00)
> Configuration.#2: len 0x05 type 0x03
> Configuration.Location: 0x00000001
> Configuration.Location: Front Left (0x00000001)
> Configuration.#3: len 0x03 type 0x04
> Configuration.Frame Length: 90 (0x005a)
> Configuration.#4: len 0x02 type 0x05
> Configuration.Frame Blocks per SDU: 1 (0x01)
> Device: /org/bluez/hci0/dev_70_5A_6F_63_B6_41
> State: idle
> Volume: 0x00c8 (200)
> Endpoint: /org/bluez/hci0/dev_70_5A_6F_63_B6_41/pac_sink0
> QoS.CIG: 0x00 (0)
> QoS.CIS: 0x00 (0)
> QoS.Framing: 0x00 (0)
> QoS.PresentationDelay: 0x00009c40 (40000)
> QoS.Interval: 0x00001d4c (7500)
> QoS.Latency: 0x000f (15)
> QoS.SDU: 0x005a (90)
> QoS.PHY: 0x02 (2)
> QoS.Retransmissions: 0x05 (5)
> Location: 0x00000003 (3)
> Links: /org/bluez/hci0/dev_70_5A_6F_63_B6_41/pac_source0/fd1
> Transport /org/bluez/hci0/dev_70_5A_6F_63_B6_41/pac_sink0/fd4
> UUID: Source PAC (00002bcb-0000-1000-8000-00805f9b34fb)
> Codec: 0x06 (6)
> Configuration.#0: len 0x02 type 0x01
> Configuration.Sampling Frequency: 48 Khz (0x08)
> Configuration.#1: len 0x02 type 0x02
> Configuration.Frame Duration: 7.5 ms (0x00)
> Configuration.#2: len 0x05 type 0x03
> Configuration.Location: 0x00000002
> Configuration.Location: Front Right (0x00000002)
> Configuration.#3: len 0x03 type 0x04
> Configuration.Frame Length: 90 (0x005a)
> Configuration.#4: len 0x02 type 0x05
> Configuration.Frame Blocks per SDU: 1 (0x01)
> Device: /org/bluez/hci0/dev_70_5A_6F_63_B6_41
> State: idle
> Volume: 0x00c8 (200)
> Endpoint: /org/bluez/hci0/dev_70_5A_6F_63_B6_41/pac_sink0
> QoS.CIG: 0x00 (0)
> QoS.CIS: 0x01 (1)
> QoS.Framing: 0x00 (0)
> QoS.PresentationDelay: 0x00009c40 (40000)
> QoS.Interval: 0x00001d4c (7500)
> QoS.Latency: 0x000f (15)
> QoS.SDU: 0x005a (90)
> QoS.PHY: 0x02 (2)
> QoS.Retransmissions: 0x05 (5)
> Location: 0x00000003 (3)
> Links: /org/bluez/hci0/dev_70_5A_6F_63_B6_41/pac_source0/fd2
> Transport /org/bluez/hci0/dev_70_5A_6F_63_B6_41/fd5
> UUID: Audio Source (0000110a-0000-1000-8000-00805f9b34fb)
> Codec: 0x02 (2)
> Media Codec: MPEG24
> Object Types: MPEG-4 AAC LC
> Frequencies: 48kHz
> Channels: 2
> Bitrate: 320000
> VBR: Yes
> Device: /org/bluez/hci0/dev_70_5A_6F_63_B6_41
> State: idle
> Delay: 0x0960 (2400)
> Volume: 0x0064 (100)
Forgot to mention, but with the above transports it seems to confuse
the gnome audio output selection, it doesn't seem to be able to mix
A2DP and BAP transports for some reason, so when I select the device
it enables BAP but A2DP is not shown as an option.
> We might need to check if other devices have such behavior, perhaps
> the headset is saving the last bearer it connected to so it tries to
> restore it or something like that, this may messes up with that, in
> the other hand the bearer can be selected even before first connecting
> so we may decide to default to bredr instead of last seen for this
> type of devices (Classic+LE Audio).
>
> Anyway there are some pieces left that I still need to implement like
> save the bearer in the storage so when the daemon is reloaded, or in
> case of reboot, it restores the last mode properly.
>
> --
> Luiz Augusto von Dentz
--
Luiz Augusto von Dentz
