On 20/12/23 16:31, glasswings wrote:
> DG2 sounds like an Intel GPU and HDMI, so it's also something I would
ignore for now.
Indeed, I can confirm now that the "DG Audio Controller Pro" device at
PCI address 04:00.0 is the Intel i950 GPU and HDMI controller and I am
able to use my screen's headphone output as a workaround for listening
and the USB webcam's mic as an input. Still like to get the built-in
analog controller at PCI address 00:1f.3 working though.
> My next steps would be
>
> * reboot into the firmware menu and poke around, just in case there's
a new option
OK, so I'm not sure that option is new, but I found a switch that allows
me to enable/disable the "Intel HD Audio Controller".
Here is the full `sudo dmesg` output after disabling it:
https://pastebin.com/rMeKuyC7
And after re-enabling it: https://pastebin.com/xLm2CjQt
Here are some differences I found noteworthy:
```
[ 0.308734] pci 0000:00:1f.3: [8086:7a50] type 00 class 0x040300
[ 0.308775] pci 0000:00:1f.3: reg 0x10: [mem
0x4201420000-0x4201423fff 64bit]
[ 0.308826] pci 0000:00:1f.3: reg 0x20: [mem
0x4201300000-0x42013fffff 64bit]
[ 0.308926] pci 0000:00:1f.3: PME# supported from D3hot D3cold
```
and
`[ 4.567449] snd_hda_intel 0000:00:1f.3: no codecs found!`
appears in the `sudo dmesg` output only when enabled.
The `sudo lspci -v` output only contains
```
00:1f.3 Audio device: Intel Corporation Device 7a50 (rev 11)
DeviceName: Onboard - Sound
Subsystem: Micro-Star International Co., Ltd. [MSI] Device 9e02
Flags: bus master, fast devsel, latency 32, IRQ 159, IOMMU group 10
Memory at 4201420000 (64-bit, non-prefetchable) [size=16K]
Memory at 4201300000 (64-bit, non-prefetchable) [size=1M]
Capabilities: [50] Power Management version 3
Capabilities: [80] Vendor Specific Information: Len=14 <?>
Capabilities: [60] MSI: Enable+ Count=1/1 Maskable- 64bit+
Kernel driver in use: snd_hda_intel
Kernel modules: snd_hda_intel, snd_sof_pci_intel_tgl
```
when enabled in the BIOS. Interestingly, `cat
/etc/modprobe.d/alsa-base.conf` shows two `snd_hda_intel` entries when
enabled but a single one when disabled, I'm guessing since the i950's
DG2 Audio Controller is also an Intel device, it shows up in both cases.
At this point, I'm suspecting a problem with the loading of the
snd_hda_codec_realtek hdaudioC1D0 driver, as the following block used to
be in my boot log and disappeared, along with my analog audio:
```
Dec 18 14:42:22 gar kernel: snd_hda_codec_realtek hdaudioC1D0:
autoconfig for ALC897: line_outs=1 (0x14/0x0/0x0/0x0/0x0) type:line
Dec 18 14:42:22 gar kernel: snd_hda_codec_realtek hdaudioC1D0:
speaker_outs=0 (0x0/0x0/0x0/0x0/0x0)
Dec 18 14:42:22 gar kernel: snd_hda_codec_realtek hdaudioC1D0:
hp_outs=1 (0x1b/0x0/0x0/0x0/0x0)
Dec 18 14:42:22 gar kernel: snd_hda_codec_realtek hdaudioC1D0: mono:
mono_out=0x0
Dec 18 14:42:22 gar kernel: snd_hda_codec_realtek hdaudioC1D0: inputs:
Dec 18 14:42:22 gar kernel: snd_hda_codec_realtek hdaudioC1D0:
Front Mic=0x19
Dec 18 14:42:22 gar kernel: snd_hda_codec_realtek hdaudioC1D0: Rear
Mic=0x18
Dec 18 14:42:22 gar kernel: snd_hda_codec_realtek hdaudioC1D0:
Line=0x1a
```
Is this what you were referring to when you said
> The big question in my mind is whether the correct Alsa module is
> being detected and loaded, or perhaps some idea of whether it's
> failing.
?
I'll look at your other comment as soon as I solve the "no codecs found"
error and still don't get any sound:
> I also wonder whether the jack task assignment changed. Many HDA
> devices can switch an analog jack between several kinds of input and
> output (line-level vs microphone with DC bias vs headphone out) and I
> could imagine those settings being changed. hdajackretask, which I
> think is part of alsa-tools-gui, is the usual way to change those
> settings.
Thanks, Christian
