Hi, The issue is reproducible on systems with an Intel Audio chip and a NVIDIA GPU with HDMI output (detected as a soundcard). When both conditions are met, PulseAudio does not detect the Intel Audio soundcard anymore and no sound at all is output. The only workaround consists in inserting garbage in /etc/pulse/daemon.conf to prevent it from loading (say, a syntax error in the file). Then sound is back again. I have NVidia blob 390.87, and I'm working with KDE. Could you please fix the issue ? Thanks a lot
Hi again, I tested with pavucontrol and the Intel HDA audio controller simply does not appear. Looks like I'm not the only one affected, see : https://unix.stackexchange.com/questions/452079/no-sound-unless-root-in-debian-buster @Debian NVIDIA Maintainers -> the issue might also come from the NVidia driver (misconfiguration), although it did not occur in Stretch w/ BPO 390.87-4 driver.
Workaround : In /etc/pulse/default.pa uncomment the following line : load-module module-alsa-sink Then everything goes back fine. Looks like there's something wrong with PA detection module.
Hi, Could reproduce this bug on an host without an NVidia GPU. Workaround is still the same. Le lun. 7 janv. 2019 à 07:41, Julien Aubin <julien.aubin@gmail.com> a écrit :
Control: tags -1 moreinfo Could you please attach a verbose log of pulseaudio when the problem occurs? Sounds like the udev module is not doing its job.
Le mer. 9 janv. 2019 à 12:33, Felipe Sateler <fsateler@debian.org> a écrit : Hi How do I get the verbose log ?
Use: systemctl --user --runtime --full edit pulseaudio Append a '-vvvv' to the ExecStart line systemctl --user restart pulseaudio And then get the log journalctl --user-unit pulseaudio --since=sometimebeforerestarting > pulse log (You might need root to get the logs) Verify that the log is not empty, and then attach it. Saludos
Le mer. 9 janv. 2019 à 12:48, Felipe Sateler <fsateler@debian.org> a écrit : Hi, Attached the logs.
The log says: module-udev-detect.c: /devices/pci0000:00/0000:00:1b.0/sound/card0 is busy: yes Some other program appears to be holding the device busy. Common culprits include espeakup and timidity. You can further inspect with: sudo lsof /dev/snd/*
Dear Maintainer, I think I'm seeing this problem. Loading module-alsa-sink makes the device appear though there's no inputs. Workarounds for bugs #901148 and #901689 don't work. My Nvidia driver version: ii nvidia-legacy-390xx-driver 390.116-1 amd64 NVIDIA metapackage (390xx legacy version) Best regards.
Sorry, the contents of the file /etc/pulse/default.pa in my last reply are incorrect. Lines: load-module module-alsa-sink load-module module-alsa-source device=hw:1,0 should be commented like this: #load-module module-alsa-sink #load-module module-alsa-source device=hw:1,0 Since that's how they were when the log was recorded.
Thank you so much for this. I had the same symptoms after upgrading to Buster. Removing timidity "solved" the issue. As a default behaviour, this is pretty bad. Is this a known issue? Should this be filed as another bug somewhere?
Control: forcemerge 901148 -1 This is already tracked as bug #901148 in timidity. I'm Merging the bugs .
Hey there, Could you help me received 4.2 million for Charity Projects in your country?