fix: Check for PipeWire as well as PulseAudio before falling back to ALSA

[alexhaydock]

Description

• Resolves bug: VMs crashing on Fedora 41 as a result of #1480 #1564

Type of change

• Bug fix (non-breaking change which fixes an issue)

Checklist:

• I have performed a self-review of my code
• I have tested my code in common scenarios and confirmed there are no regressions
• I have added comments to my code, particularly in hard-to-understand sections

[szorfein]

This will not work for pure Alsa system. Pipewire is always installed as dependencies for another app even if not used.

[szorfein]

Maybe instead check if the pipewire service is running like

systemctl is-active pipewire

[alexhaydock]

A good point, though I suppose that would have similar downsides when running Quickemu on non-systemd distros like Alpine or Void. I guess this check might need to get more complex and layered.

…

On Sun, Jan 19, 2025 at 14:37, Szorfein ***@***.***(mailto:On Sun, Jan 19, 2025 at 14:37, Szorfein <<a href=)> wrote: Maybe instead check if the pipewire service is running like systemctl is-active pipewire — Reply to this email directly, [view it on GitHub](#1565 (comment)), or [unsubscribe](https://github.com/notifications/unsubscribe-auth/ADEK7VQSRMBNHLKTGKAUJTL2LO2AZAVCNFSM6AAAAABVOTTHGGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDMMBQHA4DOOBUG4). You are receiving this because you authored the thread.Message ID: ***@***.***>

[szorfein]

if work for me if testing with pipewire-pulse if u can test

if ! command -v pacmd >/dev/null 2>&1 && ! command -v pipewire-pulse >/dev/null 2>&1; then

[alexhaydock]

if work for me if testing with pipewire-pulse if u can test

if ! command -v pacmd >/dev/null 2>&1 && ! command -v pipewire-pulse >/dev/null 2>&1; then

Thanks! Have updated the PR to look for pipewire-pulse instead of just pipewire.

I think that's definitely more robust and I can confirm it works to solve the original bug still on Fedora 41. Should also cope with situations where the system is pure ALSA still, like you mentioned before.

[lj3954]

Could we not just check whether one of these servers is actively running?

if ! pidof pipewire 2>&1 /dev/null && ! pidof pulseaudio 2>&1 /dev/null; then

[lj3954]

QEMU 8.1 added a pipewire audio backend. We should use it if possible. This should be a better solution in general.

[alexhaydock]

QEMU 8.1 added a pipewire audio backend. We should use it if possible. This should be a better solution in general.

This looks good to me. Makes sense to actually use PipeWire natively if available in both QEMU and on the host. My testing on Fedora 41 suggests this works perfectly.

Having accepted your suggestion directly within the GitHub UI, there are now two commits with different solutions to the original issue so when merging it will make sense to squash them.