hi nvidia:
我快速的连续点击左右声道测试按钮，声音会卡住，再按时就没有声音了，该如何排查问题。
版本信息：
uname -a
Linux vision 6.8.12-rt-tegra #3 SMP PREEMPT_RT Tue Jun 23 10:07:50 HKT 2026 aarch64 aarch64 aarch64 GNU/Linux
cat /etc/nv_tegra_release
TARGET_USERSPACE_LIB_DIR=nvidia
TARGET_USERSPACE_LIB_DIR_PATH=usr/lib/aarch64-linux-gnu/nvidia
— 🤖 This is an automated AI response. I’m here to help, but please verify important details! —
I could not find clear support for that in the provided Jetson AGX Orin documentation.
— We support team will also look into your issue to have further suggestions soon. please stay tuned. —
Hi Carol,
Please share the logs mentioned below before and after reproducing the issue.
hi nvidia:
any update ?
Sorry for the delay.
I couldn’t find any obvious issue from the attached logs.
Please try clicking the left/right channel test buttons with a 2-3 second pause between each click. Also try waiting until the current test sound fully finishes before clicking the next channel.
If the issue only happens with rapid repeated clicks, but not with a pause, it suggests the audio stream cleanup/reopen path is racing or getting stuck. Please share:
hi nvidia:
2. log when issue happens
$ sudo fuser -v /dev/snd/pcmCDp /dev/snd/controlC*
[sudo] password for walker:
Specified filename /dev/snd/pcmCDp does not exist.
USER PID ACCESS COMMAND
/dev/snd/controlC0: walker 2251 F.... pipewire
walker 2255 F.... wireplumber
/dev/snd/controlC1: walker 2255 F.... wireplumber
$ cat /proc/asound/card*/pcmOp/sub0/hw_params
cat: '/proc/asound/card*/pcmOp/sub0/hw_params': No such file or directory
$ cat /proc/asound/card*/pcm0p/sub0/hw_params
access: MMAP_INTERLEAVED
format: S24_LE
subformat: STD
channels: 2
rate: 48000 (48000/1)
period_size: 512
buffer_size: 4096
$ pactl list short sink-inputs
227 4294967295 198 PipeWire s16le 2ch 44100Hz
229 4294967295 101 PipeWire s16le 2ch 44100Hz
$
hi nvidia:
any update ?
Hi,
I did not find any obvious kernel-level issue from the logs shared so far.
Could you please also share the /var/log/syslog logs before and after the failure?
Also, please check cat /proc/asound/card*/pcm0p/sub0/hw_params when one channel finishes playing, and confirm whether the stream is still open or already closed.
In addition, during reproduction, please also share:
alsactl monitor
journalctl --user -f -u pipewire -u wireplumber -u pipewire-pulse -u pulseaudio
DAPM trace logs by enabling snd_soc_ trace events and collecting cat /sys/kernel/debug/tracing/trace_pipe | grep \*
If the above still looks clean, please also share regmap trace logs (regmap_reg_read / regmap_reg_write) during reproduction, since this may be an intermittent control-path issue.
For the “fast” L/R switching case, could you please clarify how exactly you are doing it? If possible, please share a short video for reference.
audio-log.zip (18.5 MB) this is audio log
Rapid Left/Right clicks in GNOME Sound test can create multiple short-lived test streams, but gnome-control-center does not always release the older streams cleanly. Those stale streams remain active in PipeWire as media.role=test sink-inputs.
Evidence:
application.process.binary = “gnome-control-center”
media.role = “test”
event.id = “audio-channel-front-left/right”
stream.is-live = “true”
Corked: no
Because PipeWire still sees these as live client-owned streams, it keeps them linked to the sink. After that, the Sound test button becomes very slow to enable/disable and may take minutes.
Check:
pactl list short sink-inputs
Steps to recover from hang state:
pkill -f '/usr/bin/gnome-control-center'
systemctl --user restart pipewire-pulse wireplumber pipewire
The issue is related to the GNOME Sound Settings channel-test path.