Internal microphone does not work with 5.12 (Tiger Lake, X1 Nano)
Hello,
it seems that the internal microphone does not work anymore with 5.12.6/5.12.7. I am somewhat sure it used to work within the 5.11.x series and that sound did not work at all on 5.10.
I have uploaded the hardware information on
http://alsa-project.org/db/?f=f1a2b32db122695b2cf736db27cb6ed2fa8a9042
The microphone of a plugged in headset however *does* work.
Is this a known bug/regression and is there anything I can do to debug it?
Best regards,
Nico
-- Sustainable and modern Infrastructures by ungleich.ch
On 27. 05. 21 11:50, Nico Schottelius wrote:
Hello,
it seems that the internal microphone does not work anymore with 5.12.6/5.12.7. I am somewhat sure it used to work within the 5.11.x series and that sound did not work at all on 5.10.
I have uploaded the hardware information on
http://alsa-project.org/db/?f=f1a2b32db122695b2cf736db27cb6ed2fa8a9042
The microphone of a plugged in headset however *does* work.
Is this a known bug/regression and is there anything I can do to debug it?
Could you check, if the Intel SOF driver is enabled in your kernel?
CONFIG_SND_SOC_SOF_TIGERLAKE
Jaroslav
Jaroslav Kysela perex@perex.cz writes:
On 27. 05. 21 11:50, Nico Schottelius wrote: Could you check, if the Intel SOF driver is enabled in your kernel?
CONFIG_SND_SOC_SOF_TIGERLAKE
Just checked, it was disabled. I rebuilt 5.12.6, rebooted, but now the sound card is gone (no device files in /dev/snd anymore), proably failing due to:
[ 55.713668] snd_hda_intel 0000:00:1f.3: DSP detected with PCI class/subclass/prog-if info 0x040380 [ 55.713679] snd_hda_intel 0000:00:1f.3: Digital mics found on Skylake+ platform, using SOF driver [ 55.745863] sof-audio-pci-intel-tgl 0000:00:1f.3: DSP detected with PCI class/subclass/prog-if info 0x040380 [ 55.745872] sof-audio-pci-intel-tgl 0000:00:1f.3: Digital mics found on Skylake+ platform, using SOF driver [ 55.746033] sof-audio-pci-intel-tgl 0000:00:1f.3: DSP detected with PCI class/subclass/prog-if 0x040380 [ 55.748315] sof-audio-pci-intel-tgl 0000:00:1f.3: use msi interrupt mode [ 55.752671] sof-audio-pci-intel-tgl 0000:00:1f.3: warning: No matching ASoC machine driver found [ 55.752673] sof-audio-pci-intel-tgl 0000:00:1f.3: error: no matching ASoC machine driver found - aborting probe [ 55.752673] sof-audio-pci-intel-tgl 0000:00:1f.3: error: failed to get machine info -19 [ 55.752764] sof-audio-pci-intel-tgl 0000:00:1f.3: error: failed to probe DSP hardware!
I have uploaded the full report on http://alsa-project.org/db/?f=d3c7ba9611b3c823709d7bce0df0f592b3471735
Nico
-- Sustainable and modern Infrastructures by ungleich.ch
Hi Nico,
On Thu, 27 May 2021, Nico Schottelius wrote:
Just checked, it was disabled. I rebuilt 5.12.6, rebooted, but now the sound card is gone (no device files in /dev/snd anymore), proably failing due to:
[ 55.713668] snd_hda_intel 0000:00:1f.3: DSP detected with PCI class/subclass/prog-if info 0x040380 [ 55.713679] snd_hda_intel 0000:00:1f.3: Digital mics found on Skylake+ platform, using SOF driver [ 55.745863] sof-audio-pci-intel-tgl 0000:00:1f.3: DSP detected with PCI class/subclass/prog-if info 0x040380 [ 55.745872] sof-audio-pci-intel-tgl 0000:00:1f.3: Digital mics found on Skylake+ platform, using SOF driver [ 55.746033] sof-audio-pci-intel-tgl 0000:00:1f.3: DSP detected with PCI class/subclass/prog-if 0x040380 [ 55.748315] sof-audio-pci-intel-tgl 0000:00:1f.3: use msi interrupt mode [ 55.752671] sof-audio-pci-intel-tgl 0000:00:1f.3: warning: No matching ASoC machine driver found [ 55.752673] sof-audio-pci-intel-tgl 0000:00:1f.3: error: no matching ASoC machine driver found - aborting probe [ 55.752673] sof-audio-pci-intel-tgl 0000:00:1f.3: error: failed to get machine info -19 [ 55.752764] sof-audio-pci-intel-tgl 0000:00:1f.3: error: failed to probe DSP hardware!
it seems more kernel config options are missing. Distribution kernels typically enable all the machine drivers, but you are specifically missing CONFIG_SND_SOC_INTEL_SKL_HDA_DSP_GENERIC_MACH=m
Distribution kernels are a good reference when building on your own, and we in SOF also maintain example kconfig settings for typical systems. Here's the bit for systems with DSP and HDA codecs: https://github.com/thesofproject/kconfig/blob/master/hdaudio-codecs-defconfi...
More info at: https://github.com/thesofproject/kconfig/
Br, Kai
Hey Kai, Kai Vehmanen kai.vehmanen@linux.intel.com writes:
Hi Nico,
On Thu, 27 May 2021, Nico Schottelius wrote:
Just checked, it was disabled. I rebuilt 5.12.6, rebooted, but now the sound card is gone (no device files in /dev/snd anymore), proably failing due to:
[ 55.713668] snd_hda_intel 0000:00:1f.3: DSP detected with PCI class/subclass/prog-if info 0x040380 [ 55.713679] snd_hda_intel 0000:00:1f.3: Digital mics found on Skylake+ platform, using SOF driver [ 55.745863] sof-audio-pci-intel-tgl 0000:00:1f.3: DSP detected with PCI class/subclass/prog-if info 0x040380 [ 55.745872] sof-audio-pci-intel-tgl 0000:00:1f.3: Digital mics found on Skylake+ platform, using SOF driver [ 55.746033] sof-audio-pci-intel-tgl 0000:00:1f.3: DSP detected with PCI class/subclass/prog-if 0x040380 [ 55.748315] sof-audio-pci-intel-tgl 0000:00:1f.3: use msi interrupt mode [ 55.752671] sof-audio-pci-intel-tgl 0000:00:1f.3: warning: No matching ASoC machine driver found [ 55.752673] sof-audio-pci-intel-tgl 0000:00:1f.3: error: no matching ASoC machine driver found - aborting probe [ 55.752673] sof-audio-pci-intel-tgl 0000:00:1f.3: error: failed to get machine info -19 [ 55.752764] sof-audio-pci-intel-tgl 0000:00:1f.3: error: failed to probe DSP hardware!
it seems more kernel config options are missing. Distribution kernels typically enable all the machine drivers, but you are specifically missing CONFIG_SND_SOC_INTEL_SKL_HDA_DSP_GENERIC_MACH=m
I've enabled this one and all machine types and indeed the card is being recognised (report at http://alsa-project.org/db/?f=5cbe7c293cbb80ab548b892cfc8b991476b0b2aa). I checked permissions on /dev/snd, which looks good.
However, obs does not list the microphone at all and chromium / jitsi shows the snd-hda-dsp mic, however does not let me select it.
So with above options:
- playback works - microphone does not work
Distribution kernels are a good reference when building on your own, and we in SOF also maintain example kconfig settings for typical systems. Here's the bit for systems with DSP and HDA codecs: https://github.com/thesofproject/kconfig/blob/master/hdaudio-codecs-defconfi...
I have used the .config from the linux-edge kernel from Alpine Linux as a basis. The above configuration looks to be enabled in my .config as well.
Alpine's default config might include a bit less options than other distros, I am currently in touch with the developers there for patching it, as soon as I got the sound card properly working.
Best,
Nico
-- Sustainable and modern Infrastructures by ungleich.ch
Hi Nico,
On Thu, 27 May 2021, Nico Schottelius wrote:
it seems more kernel config options are missing. Distribution kernels typically enable all the machine drivers, but you are specifically missing CONFIG_SND_SOC_INTEL_SKL_HDA_DSP_GENERIC_MACH=m
I've enabled this one and all machine types and indeed the card is being recognised (report at http://alsa-project.org/db/?f=5cbe7c293cbb80ab548b892cfc8b991476b0b2aa). I checked permissions on /dev/snd, which looks good.
However, obs does not list the microphone at all and chromium / jitsi shows the snd-hda-dsp mic, however does not let me select it.
ok, that's great so the driver appears to work now. You could try to do simple arecord test in terminal:
arecord -fdat -vv -Dplugw:0,6 -c4 /dev/null
You should see vumeter activity from the built-in microphone.
If that works, this starts to go a bit outside the scope of this list, but it sounds something in Alpine's user-space integration is missing. In e.g. Fedora and Ubuntu, ALSA UCM file is provided for these platforms and the shipped versions of Pipewire/Pulseaudio can expose the devices correctly to apps. You can do a sanity test (of your hardware) by booting a recent live image of either distro.
Br, Kai
Hey Kai, Kai Vehmanen kai.vehmanen@linux.intel.com writes:
Hi Nico,
On Thu, 27 May 2021, Nico Schottelius wrote:
it seems more kernel config options are missing. Distribution kernels typically enable all the machine drivers, but you are specifically missing CONFIG_SND_SOC_INTEL_SKL_HDA_DSP_GENERIC_MACH=m
I've enabled this one and all machine types and indeed the card is being recognised (report at http://alsa-project.org/db/?f=5cbe7c293cbb80ab548b892cfc8b991476b0b2aa). I checked permissions on /dev/snd, which looks good.
However, obs does not list the microphone at all and chromium / jitsi shows the snd-hda-dsp mic, however does not let me select it.
ok, that's great so the driver appears to work now. You could try to do simple arecord test in terminal:
arecord -fdat -vv -Dplugw:0,6 -c4 /dev/null
That does not work:
[15:55] nb3:~% arecord -fdat -vv -Dplugw:0,6 -c4 /dev/null ALSA lib pcm.c:2660:(snd_pcm_open_noupdate) Unknown PCM plugw:0,6 arecord: main:830: audio open error: No such file or directory
However, recording with audacity produces to results:
- no sound recorded by default - sound recorded when I plugin an external headset
And using arecord -fdat -vv -c4 /dev/null (without -Dplugw) I get output, but the level is always 00% without the headset, ranges 02 ~ 90% with the headset plugged in.
I was wondering if there's a hardware switch "broken" for the detection of the headset and that's why the mic is muted internally?
-- Sustainable and modern Infrastructures by ungleich.ch
arecord -fdat -vv -Dplugw:0,6 -c4 /dev/null
That does not work:
[15:55] nb3:~% arecord -fdat -vv -Dplugw:0,6 -c4 /dev/null ALSA lib pcm.c:2660:(snd_pcm_open_noupdate) Unknown PCM plugw:0,6 arecord: main:830: audio open error: No such file or directory
typo: you need to use plughw:0,6 or just hw:0,6
However, recording with audacity produces to results:
- no sound recorded by default
- sound recorded when I plugin an external headset
And using arecord -fdat -vv -c4 /dev/null (without -Dplugw) I get output, but the level is always 00% without the headset, ranges 02 ~ 90% with the headset plugged in.
I was wondering if there's a hardware switch "broken" for the detection of the headset and that's why the mic is muted internally?
No, the DMIC has nothing to do with the headset, it's an independent interface.
I am not sure if this is related to the MIC not working:
[ 160.122834] snd_hda_codec_realtek ehdaudio0D0: didn't find PCM for DAI Digital Codec DAI [ 160.122837] snd_hda_codec_realtek ehdaudio0D0: ASoC: error at snd_soc_dai_startup on Digital Codec DAI: -22 [ 160.122839] Digital Playback and Capture: ASoC: BE open failed -22 [ 160.122841] HDA Digital: ASoC: failed to start some BEs -22 [ 160.122881] snd_hda_codec_realtek ehdaudio0D0: didn't find PCM for DAI Digital Codec DAI [ 160.122882] snd_hda_codec_realtek ehdaudio0D0: ASoC: error at snd_soc_dai_startup on Digital Codec DAI: -22 [ 160.122883] Digital Playback and Capture: ASoC: BE open failed -22 [ 160.122884] HDA Digital: ASoC: failed to start some BEs -22
I posted the full output on http://alsa-project.org/db/?f=45fc657ecf37976734035202625a3c08841d7361
Firmware wise I downloaded the sof firmware from https://github.com/thesofproject/sof-bin/, I am not sure if the "unknown sof_ext_man header" message could be the problem:
[ 48.431640] sof-audio-pci-intel-tgl 0000:00:1f.3: Firmware info: version 1:7:0-47d07 [ 48.431642] sof-audio-pci-intel-tgl 0000:00:1f.3: Firmware: ABI 3:18:1 Kernel ABI 3:18:0 [ 48.431643] sof-audio-pci-intel-tgl 0000:00:1f.3: unknown sof_ext_man header type 3 size 0x30 [ 48.530671] sof-audio-pci-intel-tgl 0000:00:1f.3: Firmware info: version 1:7:0-47d07 [ 48.530675] sof-audio-pci-intel-tgl 0000:00:1f.3: Firmware: ABI 3:18:1 Kernel ABI 3:18:0 [ 48.534529] sof-audio-pci-intel-tgl 0000:00:1f.3: Topology: ABI 3:18:1 Kernel ABI 3:18:0
I also verified: the microphone of an external headset still works - so basically the behaviour is the same as without the sof module.
Nico Schottelius nico.schottelius@ungleich.ch writes:
Hey Kai, Kai Vehmanen kai.vehmanen@linux.intel.com writes:
Hi Nico,
On Thu, 27 May 2021, Nico Schottelius wrote:
Just checked, it was disabled. I rebuilt 5.12.6, rebooted, but now the sound card is gone (no device files in /dev/snd anymore), proably failing due to:
[ 55.713668] snd_hda_intel 0000:00:1f.3: DSP detected with PCI class/subclass/prog-if info 0x040380 [ 55.713679] snd_hda_intel 0000:00:1f.3: Digital mics found on Skylake+ platform, using SOF driver [ 55.745863] sof-audio-pci-intel-tgl 0000:00:1f.3: DSP detected with PCI class/subclass/prog-if info 0x040380 [ 55.745872] sof-audio-pci-intel-tgl 0000:00:1f.3: Digital mics found on Skylake+ platform, using SOF driver [ 55.746033] sof-audio-pci-intel-tgl 0000:00:1f.3: DSP detected with PCI class/subclass/prog-if 0x040380 [ 55.748315] sof-audio-pci-intel-tgl 0000:00:1f.3: use msi interrupt mode [ 55.752671] sof-audio-pci-intel-tgl 0000:00:1f.3: warning: No matching ASoC machine driver found [ 55.752673] sof-audio-pci-intel-tgl 0000:00:1f.3: error: no matching ASoC machine driver found - aborting probe [ 55.752673] sof-audio-pci-intel-tgl 0000:00:1f.3: error: failed to get machine info -19 [ 55.752764] sof-audio-pci-intel-tgl 0000:00:1f.3: error: failed to probe DSP hardware!
it seems more kernel config options are missing. Distribution kernels typically enable all the machine drivers, but you are specifically missing CONFIG_SND_SOC_INTEL_SKL_HDA_DSP_GENERIC_MACH=m
I've enabled this one and all machine types and indeed the card is being recognised (report at http://alsa-project.org/db/?f=5cbe7c293cbb80ab548b892cfc8b991476b0b2aa). I checked permissions on /dev/snd, which looks good.
However, obs does not list the microphone at all and chromium / jitsi shows the snd-hda-dsp mic, however does not let me select it.
So with above options:
- playback works
- microphone does not work
Distribution kernels are a good reference when building on your own, and we in SOF also maintain example kconfig settings for typical systems. Here's the bit for systems with DSP and HDA codecs: https://github.com/thesofproject/kconfig/blob/master/hdaudio-codecs-defconfi...
I have used the .config from the linux-edge kernel from Alpine Linux as a basis. The above configuration looks to be enabled in my .config as well.
Alpine's default config might include a bit less options than other distros, I am currently in touch with the developers there for patching it, as soon as I got the sound card properly working.
Best,
Nico
-- Sustainable and modern Infrastructures by ungleich.ch
participants (4)
-
Jaroslav Kysela -
Kai Vehmanen -
Nico Schottelius -
Pierre-Louis Bossart