bluetoothd: connect/disconnect - "AVRCP PDU parameters length don't match"

[^rooker]

I'm trying to connect to a bluetooth speaker.

The device is found, I can pair and connect - the speaker even (literally) "says": 'connected', and /var/syslog confirms that. All looks great; pulseaudio switches to the new audio sink - then "connected: no".

/var/syslog also contains *a lot of messages* regarding that connection (will post separately).

UPDATE: After adding "auto_switch=false" to my pulseaudio where it auto-loads the "bluetooth-policy" module, then `sysctl restart bluetooth` and `pulseaudio -k`: and after some time I *could finally connect* to this (and other) bluetooth speaker devices that just would never stay connected, but reconnect or bail out all the time.

I have no idea what I have done to fix this.
Basically, I copy/pasted from: https://askubuntu.com/questions/1205749/how-permanently-remove-or-disable-hsp-hfp-bluetooth-profile

I don't know yet, if this is reboot safe.
I dare to try, before I haven't understood what happened here.

I'm grateful for any insights or tips to get this "fixed" upstream?
Thank you.

Here's what I've modified in `/etc/pulse/default.pa`:

```
### Automatically load driver modules for Bluetooth hardware
.ifexists module-bluetooth-policy.so
#ORG: load-module module-bluetooth-policy
load-module module-bluetooth-policy auto_switch=false
.endif
```

--
I'm running Xubuntu 20.04
bluez 5.53-0ubuntu3.7
pulseaudio 1:13.99.1-1ubuntu3.13

[^rooker (rooker)]

Sorry for the noise, but this info may be useful?
(I've put my bluetoothd into debug verbosity)

Below are parts and the full copy of the /var/syslog part when trying to connect, then disconnecting.

The first part looks really good (IMO).
Then "something" happens - and *disconnected*.

# Highlights (summary):

```
SINK_STATE_DISCONNECTED -> SINK_STATE_CONNECTING
AVDTP: connected signaling channel to 40:EF:4C:C5:1B:6C
DISCOVER request succeeded
SET_CONFIGURATION request succeeded
TRANSPORT_STATE_REQUESTING -> TRANSPORT_STATE_ACTIVE
SINK_STATE_CONNECTED -> SINK_STATE_PLAYING
AVCTP Connected
```

**...followed immediately by:**

```
40:EF:4C:C5:1B:6C: No service update

bluetoothd[1000416]: profiles/audio/avrcp.c:handle_vendordep_pdu() AVRCP PDU 0x10, company 0x001958 len 0x0001
bluetoothd[1000416]: profiles/audio/avrcp.c:handle_vendordep_pdu() AVRCP PDU parameters length don't match
bluetoothd[1000416]: profiles/audio/avrcp.c:handle_vendordep_pdu() AVRCP PDU 0x30, company 0x001958 len 0x0000
```

**And then it all goes down:**

```
Device 40:EF:4C:C5:1B:6C disconnected, reason 0
STREAMING -> IDLE
TRANSPORT_STATE_ACTIVE -> TRANSPORT_STATE_IDLE
connected -> disconnected (0)
disconnecting -> disconnected (0)
AVCTP: closing uinput for 40:EF:4C:C5:1B:6C
acpid: input device has been disconnected, fd 26
SINK_STATE_PLAYING -> SINK_STATE_DISCONNECTED
```

Here's the full log:
=============================

```
Jan 29 19:31:13 silverstar bluetoothd[1000416]: src/device.c:connect_profiles() /org/bluez/hci0/dev_40_EF_4C_C5_1B_6C (all), client :1.291165
Jan 29 19:31:13 silverstar bluetoothd[1000416]: profiles/audio/a2dp.c:a2dp_sink_connect() path /org/bluez/hci0/dev_40_EF_4C_C5_1B_6C
Jan 29 19:31:13 silverstar bluetoothd[1000416]: profiles/audio/avdtp.c:avdtp_ref() 0x562e5c523e10: ref=1
Jan 29 19:31:13 silverstar bluetoothd[1000416]: profiles/audio/avdtp.c:avdtp_ref() 0x562e5c523e10: ref=2
Jan 29 19:31:13 silverstar bluetoothd[1000416]: profiles/audio/a2dp.c:setup_ref() 0x562e5c511f50: ref=1
Jan 29 19:31:13 silverstar bluetoothd[1000416]: profiles/audio/sink.c:sink_set_state() State changed /org/bluez/hci0/dev_40_EF_4C_C5_1B_6C: SINK_STATE_DISCONNECTED -> SINK_STATE_CONNECTING
Jan 29 19:31:13 silverstar bluetoothd[1000416]: profiles/audio/sink.c:sink_connect() stream creation in progress
Jan 29 19:31:13 silverstar bluetoothd[1000416]: src/service.c:change_state() 0x562e5c4ff310: device 40:EF:4C:C5:1B:6C profile a2dp-sink state changed: disconnected -> connecting (0)
Jan 29 19:31:14 silverstar bluetoothd[1000416]: src/adapter.c:connected_callback() hci0 device 40:EF:4C:C5:1B:6C connected eir_len 24
Jan 29 19:31:14 silverstar bluetoothd[1000416]: profiles/audio/avdtp.c:avdtp_connect_cb() AVDTP: connected signaling channel to 40:EF:4C:C5:1B:6C
Jan 29 19:31:14 silverstar bluetoothd[1000416]: profiles/audio/avdtp.c:avdtp_connect_cb() AVDTP imtu=672, omtu=672
Jan 29 19:31:14 silverstar bluetoothd[1000416]: profiles/audio/avdtp.c:avdtp_register_remote_sep() seid 1 type 1 media 0
Jan 29 19:31:14 silverstar bluetoothd[1000416]: profiles/audio/a2dp.c:register_remote_sep() Found remote SEP: /org/bluez/hci0/dev_40_EF_4C_C5_1B_6C/sep1
Jan 29 19:31:15 silverstar blue...

[^rooker (rooker)]

I found something!
Even with the above mentioned "patch", the connection-behavior was still behaving erratic.

> IT NEEDED MUSIC! :D

Seriously: I happened to have Clementine music player running - playing music *while connecting* to all of these (previously problematic) BT-speakers:

*all of them* work when I play music while connecting.
Awesome.

However: What is the problem with "empty/silent buffers"?