Created attachment 289135 [details] screen from alsamixer when not working Sometimes "HD-Audio Generic Headphone" device is not detected during kernel boot. As a result, headphones do work alongside with built-in speakers and it's not possible to enable auto-mute in alsamixer. Attached screen from alsamixer. The bug is not 100% reproducible, reboot usually helps. DMESG when not working: [ 19.191827] input: HD-Audio Generic HDMI/DP,pcm=3 as /devices/pci0000:00/0000:00:08.1/0000:05:00.1/sound/card0/input10 [ 19.192055] input: HD-Audio Generic HDMI/DP,pcm=7 as /devices/pci0000:00/0000:00:08.1/0000:05:00.1/sound/card0/input11 [ 19.192249] input: HD-Audio Generic HDMI/DP,pcm=8 as /devices/pci0000:00/0000:00:08.1/0000:05:00.1/sound/card0/input12 [ 19.199418] snd_hda_codec_conexant hdaudioC1D0: CX8070: BIOS auto-probing. [ 19.200336] snd_hda_codec_conexant hdaudioC1D0: autoconfig for CX8070: line_outs=1 (0x17/0x0/0x0/0x0/0x0) type:speaker [ 19.200338] snd_hda_codec_conexant hdaudioC1D0: speaker_outs=0 (0x0/0x0/0x0/0x0/0x0) [ 19.200340] snd_hda_codec_conexant hdaudioC1D0: hp_outs=0 (0x0/0x0/0x0/0x0/0x0) [ 19.200341] snd_hda_codec_conexant hdaudioC1D0: mono: mono_out=0x0 [ 19.200342] snd_hda_codec_conexant hdaudioC1D0: inputs: [ 19.200345] snd_hda_codec_conexant hdaudioC1D0: Internal Mic=0x1a [ 19.200347] snd_hda_codec_conexant hdaudioC1D0: Mic=0x19 [ 19.209160] input: HD-Audio Generic Mic as /devices/pci0000:00/0000:00:08.1/0000:05:00.6/sound/card1/input13 [ 19.226055] input: PC Speaker as /devices/platform/pcspkr/input/input14 DMESG when working [ 31.659625] input: HD-Audio Generic HDMI/DP,pcm=3 as /devices/pci0000:00/0000:00:08.1/0000:05:00.1/sound/card0/input10 [ 31.659742] input: HD-Audio Generic HDMI/DP,pcm=7 as /devices/pci0000:00/0000:00:08.1/0000:05:00.1/sound/card0/input11 [ 31.659806] input: HD-Audio Generic HDMI/DP,pcm=8 as /devices/pci0000:00/0000:00:08.1/0000:05:00.1/sound/card0/input12 [ 31.680364] snd_hda_codec_conexant hdaudioC1D0: CX8070: BIOS auto-probing. [ 31.681298] snd_hda_codec_conexant hdaudioC1D0: autoconfig for CX8070: line_outs=1 (0x17/0x0/0x0/0x0/0x0) type:speaker [ 31.681301] snd_hda_codec_conexant hdaudioC1D0: speaker_outs=0 (0x0/0x0/0x0/0x0/0x0) [ 31.681302] snd_hda_codec_conexant hdaudioC1D0: hp_outs=1 (0x16/0x0/0x0/0x0/0x0) [ 31.681303] snd_hda_codec_conexant hdaudioC1D0: mono: mono_out=0x0 [ 31.681304] snd_hda_codec_conexant hdaudioC1D0: inputs: [ 31.681306] snd_hda_codec_conexant hdaudioC1D0: Internal Mic=0x1a [ 31.681307] snd_hda_codec_conexant hdaudioC1D0: Mic=0x19 [ 31.702621] input: HD-Audio Generic Mic as /devices/pci0000:00/0000:00:08.1/0000:05:00.6/sound/card1/input13 [ 31.702674] input: HD-Audio Generic Headphone as /devices/pci0000:00/0000:00:08.1/0000:05:00.6/sound/card1/input14
Created attachment 289137 [details] alsamixer when working
Created attachment 289139 [details] alsa mixer when working
Do you see any relevant kernel messages or such if the problem happens? i.e. is it some codec communication error? Unless we see it as an error, it's hard to debug, unfortunately. In anyway, please give alsa-info.sh output. Run the script with --no-upload option and attach to Bugzilla. A best, give the outputs on both working and non-working cases.
Created attachment 289757 [details] Log from alsa-info.sh when everything is OK
Created attachment 289759 [details] Log from alsa-info.sh when failing
Thanks for looking at this! Sorry for the delay, I just attached the logs from alsa-info.sh. In the meantime I've found a 100% repro for this: the problem occurs always when I reboot computer from Windows to Linux. If I reboot from Linux to Linux it's always ok. I know that's weird, but it's 100% reproducible.
It looks like that the BIOS doesn't initialize the pin 0x16 properly when rebooted from Windows. Maybe Windows turn off the power of the pin, and BIOS skips the pin initialization. It's a bit messy to restore the behavior. Could you try the following patch?
Created attachment 289783 [details] Test fix patch
Just tested, it's perfect, thank you!
OK, I'll submit and merge the fix to upstream. It'll be backported to stable kernels once after it landed in Linus tree.
Oh, sorry for the mess. Yesterday I was checking it late at night and I didn't chceck the audio, only the "mute" led that was not working before the bug. Now, the "mute" LED works ok, but there's no sound at all, so it's even worse. .
Reopening due to the previous comment
Just to clarify - there's now a silent noise. I looks like a single sound that "hanged" and is played in a loop.
Hrm, then it's difficult to get through. Basically it's a BIOS bug that doesn't initialize the stuff, and hard to know what's still missing behind the scene...
Just to add another report - I experience this bug on my E495 too. If I boot into Windows, then reboot to Linux without shutting down first, the headphones don't work. It works fine when I boot to Linux from powered off. Like pasternak.t, I get the same hp_outs lines in dmesg: working: [ 3.212571] snd_hda_codec_conexant hdaudioC1D0: hp_outs=1 (0x16/0x0/0x0/0x0/0x0) not working: [ 3.891088] snd_hda_codec_conexant hdaudioC1D0: hp_outs=0 (0x0/0x0/0x0/0x0/0x0)
@bcampbell good to know :) Maybe we should submit this bug to Lenovo?
@pasternak.t i am probably affected too (209761), so i am willing to do what's neccessary in order to have this bug patched out. Is forums.lenovo.com apropriate place to fill a bug report to hardware vendor?
I have no idea
Sorry for digging out this quite old bug, but I observed exactly the same issue as described in comment #15. I have Lenovo ThinkPad E16 and kernel 6.1 (from Debian stable).