Dear Maintainer,

I have a bluetooth headeset that until recently worked fine (both A2DP and as Headset).
After a recent update the headset part is broken -- no sound is to be heard.

Previously, the profiles shown by
$ pactl list
where just
 headset_head_unit
 a2dp_sink
 off
all three were marked as
 available: true
and togling between them presented no difficulties.

After the update, there's additionally
 handsfree_head_unit
but now
 headset_head_unit
is always marked as
 available: no

While
 a2dp_sink
still plays audio,
switching to
 headset_head_unit
is not possible anymore, and
 handsfree_head_unit
is mute.

pavucontrol shows activity, audio level is at about 80%, but no sound comes from the headset anymore.

Switched to pipewire as per Wiki, problem persists.
In headset mode, both output and input are muted and there's no way to
change that -- output and input appear as working devices though, but
only silence is recorded and played.

After some research it seems to do with a change to the bluetooth
handling in the kernel and affects a lot of people.

Back to pulseaudio and following

 > https://gitlab.freedesktop.org/pulseaudio/pulseaudio/-/issues/1247

first answer by Igor Kovalenko
I edited

/etc/pulse/default.pa

to match

load-module module-bluetooth-discover enable_native_hfp_hf=false

and with that parameter (enable_native_hfp_hf=false) things are back to
normal.
Couldn't find out what the equivalent setting for pipewire would be,
that's why I went back to pulse.

Positive side effect: ofono seems not to be necessary anymore.

Hi,

I'm not sure what to do about this issue.

This suggest the problem is a kernel problem, but is uncovered by the new
HFP profile in pulse.


This is the new feature :)

 > Hi,
 >
 > I'm not sure what to do about this issue.

Well, to me it seems pulseaudio 15 should not not be promoted to testing
until this is sorted out -- given the current boom in video conferencing
due to working from home, a suddenly broken headset will cause a lot of
trouble. (Thankfully, I did not have to work this week so I could spend
three days figuring out what's wrong, couldn't afford this in a work week.)

AFAIU the issue is caused by a change in the btusb kernel driver and
thus affects most bluetooth devices connected via USB, fixed only in 5.13.
So either the fix needs to be back-ported to the current kernel or
pulseaudio needs to include the workaround outlined in the linked report.
At the very, very least users should be warned about a potential
breakage and how to work around that.

It's the silver lining in a deluge ... ;)

I don't know about most. Mine did not have this problem.

So either the fix needs to be back-ported to the current kernel or

By I don't know what to do, I mean that this is strictly speaking not a bug
in pulseaudio. The workaround you applied just disables the relevant
functionality.