[SOLVED] Soundcard not detected / Audio not working, new kernel.

kalenux81 · 2025-12-15 09:09:13

Hi everyone! My first post here in the forum

I wanted to ask help for this weird audio problem here many times since I've been fighting with it on and off for at least a month I've been browsing the forums and googling like crazy, trying out all sorts of fixes without solution... But now I finally got it solved so I share it!

Okay so I have quite old laptop, Acer Aspire 4820, which has integrated Intel/Realtek audio that for some reason didn't work in Linux (Arch nor Mint) with recent kernels but worked fine in Windows 10 automatically (so I know it's a driver/firmware issue). The error I got during boot was:

snd_hda_intel: 0000:00:1b.0:0 CORB reset timeout #2
hdaudio hdaudioC0D0: no AFG or MFG node found
snd_hda_intel: 0000:00:1b.0: no codecs initialized

I had alsa installed but it didn't detect any soundcard. I tried so many fixes, installed sof-firmware, all sorts of hacks in /etc/modprobe/intel-fix.conf, kernel boot parameters... compiled even kernel from source with all possible sound settings enabled that I could think of... but nothing helped!

Then I tried Linux Mint just for fun, but same problem: no audio. Also Mint was so slow, even slower than Windows 10. I already gave up, but then I was able to see some additional error message during boot with Mint that gave me little bit of hope to try once more. So I did a fresh Arch install again (to do some modprobe hack) and decided to try install linux-lts kernel... and holy cow just with linux-lts kernel the soundcard was automatically detected and working!

So eventually that's the solution: linux-lts kernel!

Now I'm running 6.12.61-1-lts which has EOL in dec 2026, so I hope nothing changes so much with the updates that it stops working again. So I'm glad now that my sound is working but still somewhat curious what might be the root problem why the other kernels have problem with the soundcard... Anyway wanted to share this solution for others tackling with sound problems with older hardware. But if somebody knows what might cause the problem feel free to say

xerxes_ · 2025-12-15 21:08:32

Don't be so much happy because this is NOT a solution: in the beginning of new year kernel 6.18 will be new lts kernel and Arch Linux use latest lts kernel and will switch to it, so you will hear silence.

To really solve your problem you should bisect kernel (or ask people here to help you with that) to find breaking change and report it to upstream, so in newer kernels sound would work for you.

Last edited by xerxes_ (2025-12-15 21:10:00)

kalenux81 · 2025-12-16 07:47:29

Oh man... how disappointing.. thanks a lot! just kidding But yes nice to know, cheers.

I guess I'm just gonna stick with the old LTS kernel since my laptop is so old and even the hardware has physical lifespan and this might break anyday.

I think they changed something in the kernel for a reason rather than being a bug and this old piece of hardware (that hardly anybody uses anymore anyways) just runs into conflict. Or maybe in the future kernel updates they change something back and it starts working again... It's been like this always with updates, support for new stuff conflicts with old stuff.. constant struggle

seth · 2025-12-16 09:11:29

You could test some kernels from https://archive.archlinux.org/packages/l/linux/ rather than blogging that you managed to get an undisclosed chip to work with a kernel that has a shelf life lower than an opened box of milk…

kalenux81 · 2025-12-16 10:20:48

Well thanks for the tip, tried with 6.18 and 6.18.1 (which are the upcoming ones) but no luck.

I blogged about my success because as far I know the lts kernel had EOL dec 2026 (https://www.kernel.org/category/releases.html) so it is atleast good to use for a year. Just read that arch will automatically switch to 6.18...

Anyways in the other threads a similar problem I didn't find any working fixes and some others didn't aswell. So atleast this one worked, hope it didn't offend you.

seth · 2025-12-16 10:24:25

This isn't about offending but looking for a sustainable solution that might also benefit others.
The plan would have been not to hope that it magically fixed w/ the upcoming kernel but to determine where things broke between 6.12 and now eg. test some 6.15 kernel and then 6.13 or 6.16 to narrow down on the culprit.

kalenux81 · 2025-12-16 10:41:09

Agreed. I'm all for sustainable solution that's why I'm even playing around with linux in a old crappy laptop To clarify more this sound problem doesn't seem to be so common so I assume it's only a problem with some older chipsets. And some have reported it's been on and off working with kernel releases. So maybe not worth of putting much energy into fixing the kernel code if the chipset is the problematic one as kernel/code naturally advances.

loqs · 2025-12-16 14:13:39

Why make such assumptions when you can determine which commit caused the breakage, read what the commit is intended to do and see if it was expected to cause this issue and if not report an issue upstream?

kalenux81 · 2025-12-16 15:07:36

Well first of all have no experience in bisecting kernel and how it's done properly to be reported upstream, or could I understand the code enough to know if it's expected to cause this. Secondly after investigating this issue yes I assume it's not a bug in kernel since I only saw a few reported issues of "disappearing legacy soundcard + new kernel", so don't have so much motivation to use my energy on this and I think the kernel developers are fine too. Rather I think it's just this chipset (more than 15 years old) conflicting with modern kernel, or could be a synergy problem of two pieces of hardware that exist in my laptop... who knows...

Anyway this works fine now with lts 6.12. which is good for a year more to go.

seth · 2025-12-16 15:18:01

You don't have to do any of that for now.
"Bisecting" for you atm just means to hone in on the version change that introduces the break, using the precompiled kernels from the ALA

Anyway this works fine now with lts 6.12. which is good for a year more to go.

=> https://www.debian.org/releases/stable/ might be a better distro for you.

Scimmia · 2025-12-16 15:22:27

kalenux81 wrote:

Anyway this works fine now with lts 6.12. which is good for a year more to go.

But will only be in the Arch repos for a month or so.

kalenux81 · 2025-12-16 15:40:25

Yes I could try all the kernels from 6.12. and find out the version where it breaks but anyhow I should update and use the newest kernel (for security reasons) or an older kernel that is supported for a longer time (lts). But since the new kernel don't work and lts does. But I understand your point if it was something to be reported to be fixed but I strongly think it's just this specific hardware that I have since there's not much talk about it.

Yes Debian has strong support. I'm playing around with Arch because I can make really minimal and still functional install with it that runs quite well even with a piece of crap laptop. And it's a nice and sustainable thought since Win10 pulled the plug also for old computers. But yea hardware and software goes hand in hand and sometimes hardware just gets obsolete even though it might still work.

kalenux81 · 2025-12-16 15:42:53

@Scimmia but I can still install it from the Arch archives? If it's lts with eol dec '26 shouldnt it get updates until then or am I wrong?

Edit: oh yea so the updates for 6.12. won't be available in Arch repos but only 6.18 lts. Anyway support for 6.12. should still be available from official kernel.org.

Last edited by kalenux81 (2025-12-16 15:49:26)

seth · 2025-12-16 15:51:27

Yes I could try all the kernels from 6.12

You could also not do that.
https://en.wikipedia.org/wiki/Bisection_method has nothing to do with compilers and because of the logarithmic scale you'll probably have to test only 5 or 6 kernels.

kalenux81 · 2025-12-16 16:02:25

seth wrote:

Yes I could try all the kernels from 6.12
You could also not do that.
https://en.wikipedia.org/wiki/Bisection_method has nothing to do with compilers and because of the logarithmic scale you'll probably have to test only 5 or 6 kernels.

Ok, good to know. But yea this lts already works, and it's either this or the newest kernel anything in between is pointless to use since I'm not reporting it to be fixed.

loqs · 2025-12-16 16:36:55

kalenux81 wrote:

I'm not reporting it to be fixed.

If gromit built the bisection kernels for you meaning you would need to only install, test and report back and then gromit reported the bug upstream to kernel.org for you would that help?

Last edited by loqs (2025-12-16 16:42:46)

kalenux81 · 2025-12-16 17:02:24

loqs wrote:

kalenux81 wrote:
I'm not reporting it to be fixed.
If gromit built the bisection kernels for you meaning you would need to only install, test and report back and then gromit reported the bug upstream to kernel.org for you would that help?

Well if you think it's important for bug upstream to know about this and somebody really want to dig into it for me it's really not so important since it already works and I think there's not many other "victims", or even with similar old and likely soon garbage hardware, and I have doubts it's even a bug but just a sad conflict with new way of code in the kernel.

Btw. who is gromit? sorry

Edit: plus it's been already over a month for me since this problem so I'm sure it would be be reported already or much more talk at least. Just FYI.

Last edited by kalenux81 (2025-12-16 17:10:04)

loqs · 2025-12-16 17:19:42

gromit is a package maintainer so the kernels he builds for you would be signed by a trusted key. I can look for upstream bugs for you but need at least the Vendor ID and Product ID of the sound card. If the card is PCI(E) based you can get the information from lspci -nnk for instance on my system the sound card entry is:

0c:00.4 Audio device [0403]: Advanced Micro Devices, Inc. [AMD] Starship/Matisse HD Audio Controller [1022:1487]
	Subsystem: ASUSTeK Computer Inc. Device [1043:87c4]
	Kernel driver in use: snd_hda_intel
	Kernel modules: snd_hda_intel

Can you also test linux 6.14 (if you do not use DKMS skip installing the headers):

pacman -U https://archive.archlinux.org/packages/l/linux/linux-6.14.arch1-1-x86_64.pkg.tar.zst https://archive.archlinux.org/packages/l/linux-headers/linux-headers-6.14.arch1-1-x86_64.pkg.tar.zst

So the VID:PID would be 1022:1487 subsystem 1043:87c4.

kalenux81 · 2025-12-16 17:24:05

loqs wrote:

gromit is a package maintainer so the kernels he builds for you would be signed by a trusted key. I can look for upstream bugs for you but need at least the Vendor ID and Product ID of the sound card. If the card is PCI(E) based you can get the information from lspci -nnk for instance on my system the sound card entry is:
0c:00.4 Audio device [0403]: Advanced Micro Devices, Inc. [AMD] Starship/Matisse HD Audio Controller [1022:1487]
	Subsystem: ASUSTeK Computer Inc. Device [1043:87c4]
	Kernel driver in use: snd_hda_intel
	Kernel modules: snd_hda_intel
Can you also test linux 6.14 (if you do not use DKMS skip installing the headers):
pacman -U https://archive.archlinux.org/packages/l/linux/linux-6.14.arch1-1-x86_64.pkg.tar.zst https://archive.archlinux.org/packages/l/linux-headers/linux-headers-6.14.arch1-1-x86_64.pkg.tar.zst
So the VID:PID would be 1022:1487 subsystem 1043:87c4.

Okay thanks, I try out that kernel. Meanwhile my soundcard info:

00:1b.0 Audio device [0403]: Intel Corporation 5 Series/3400 Series Chipset High Definition Audio [8086:3b56] (rev 05)
        Subsystem: Acer Incorporated [ALI] Device [1025:0359]
        Kernel driver in use: snd_hda_intel
        Kernel modules: snd_hda_intel

Edit: Just tested 6.14 and the sound works so the problem comes after this version. I give 6.15 a go aswell

Last edited by kalenux81 (2025-12-16 17:31:02)

kalenux81 · 2025-12-16 17:46:39

Okay after some kernel testing 6.15.9 is the last one that worked. In 6.16 sound card is not working anymore.

Edit: Just googled what changes 6.16 did and one was about USB audio offload. One article said:
"The USB updates for Linux 6.16 also cleared out some very old driver code to lighten the kernel by some 11k lines of code, so it balances things out with all of the new USB audio code merged."

So might this be the case, removed driver audio codes?

Last edited by kalenux81 (2025-12-16 17:57:28)

loqs · 2025-12-16 18:05:59

Gromit already has some bisection kernels built for 6.15-6.16:

$ git bisect start
status: waiting for both good and bad commits
$ git bisect bad v6.16
status: waiting for good commit(s), bad commit known
$ git bisect good v6.15
Bisecting: 7873 revisions left to test after this (roughly 13 steps)
[43db1111073049220381944af4a3b8a5400eda71] Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
$ git describe 
v6.15-8050-g43db11110730

So the first bisection point to test is:

# pacman -U https://pkgbuild.com/\~gromit/linux-bisection-kernels/linux-mainline-v6.15.r8050.g43db111-1-x86_64.pkg.tar.zst

Last edited by loqs (2025-12-16 18:06:38)

kalenux81 · 2025-12-16 18:11:59

loqs wrote:

Gromit already has some bisection kernels built for 6.15-6.16:

$ git bisect start
status: waiting for both good and bad commits
$ git bisect bad v6.16
status: waiting for good commit(s), bad commit known
$ git bisect good v6.15
Bisecting: 7873 revisions left to test after this (roughly 13 steps)
[43db1111073049220381944af4a3b8a5400eda71] Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
$ git describe 
v6.15-8050-g43db11110730

So the first bisection point to test is:

# pacman -U https://pkgbuild.com/\~gromit/linux-bisection-kernels/linux-mainline-v6.15.r8050.g43db111-1-x86_64.pkg.tar.zst

Sound card works with this one!

But about what I wrote earlier, if the old audio drivers were intentionally cleared out from 6.16. then it's not necessary to do more testing I think

Last edited by kalenux81 (2025-12-16 18:14:28)

loqs · 2025-12-16 18:39:24

kalenux81 wrote:

But about what I wrote earlier, if the old audio drivers were intentionally cleared out from 6.16. then it's not necessary to do more testing I think

00:1b.0 Audio device [0403]: Intel Corporation 5 Series/3400 Series Chipset High Definition Audio [8086:3b56] (rev 05)

Is a PCIE device not USB so I do not think that should be the cause of your issue.

$ git bisect good
Bisecting: 3942 revisions left to test after this (roughly 12 steps)
[11fcf368506d347088e613edf6cd2604d70c454f] uapi: bitops: use UAPI-safe variant of BITS_PER_LONG again
$ git describe 
v6.15-11981-g11fcf368506d

# pacman -U https://pkgbuild.com/\~gromit/linux-bisection-kernels/linux-mainline-headers-v6.15.r11981.g11fcf36-1-x86_64.pkg.tar.zst

kalenux81 · 2025-12-16 18:51:39

loqs wrote:

kalenux81 wrote:
But about what I wrote earlier, if the old audio drivers were intentionally cleared out from 6.16. then it's not necessary to do more testing I think
00:1b.0 Audio device [0403]: Intel Corporation 5 Series/3400 Series Chipset High Definition Audio [8086:3b56] (rev 05)
Is a PCIE device not USB so I do not think that should be the cause of your issue.
$ git bisect good
Bisecting: 3942 revisions left to test after this (roughly 12 steps)
[11fcf368506d347088e613edf6cd2604d70c454f] uapi: bitops: use UAPI-safe variant of BITS_PER_LONG again
$ git describe 
v6.15-11981-g11fcf368506d
# pacman -U https://pkgbuild.com/\~gromit/linux-bisection-kernels/linux-mainline-headers-v6.15.r11981.g11fcf36-1-x86_64.pkg.tar.zst

This one works aswell!

Is a PCIE device not USB so I do not think that should be the cause of your issue.

Yea it's a PCIE device but also the article didn't specify what kind of "old drivers" were cleared out to make room for the USB drivers. Don't know where to check but to save time might be good to make sure if the removed drivers are the cause.

loqs · 2025-12-16 19:26:01

The removed drivers were USB https://lore.kernel.org/lkml/aEKpczs8HI … kroah.com/

$ git bisect good
Bisecting: 1971 revisions left to test after this (roughly 11 steps)
[601dddb6c5d6bf63c63b2efba98231db5f861696] Merge tag 'regulator-fix-v6.16-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/regulator
$ git describe 
v6.16-rc1-147-g601dddb6c5d6

pacman -U https:/\pkgbuild.com/\~gromit/linux-bisection-kernels/linux-mainline-v6.16.rc1.r147.g601dddb-1-x86_64.pkg.tar.zst

Arch Linux

#1 2025-12-15 09:09:13

[SOLVED] Soundcard not detected / Audio not working, new kernel.

#2 2025-12-15 21:08:32

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#3 2025-12-16 07:47:29

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#4 2025-12-16 09:11:29

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#5 2025-12-16 10:20:48

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#6 2025-12-16 10:24:25

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#7 2025-12-16 10:41:09

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#8 2025-12-16 14:13:39

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#9 2025-12-16 15:07:36

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#10 2025-12-16 15:18:01

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#11 2025-12-16 15:22:27

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#12 2025-12-16 15:40:25

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#13 2025-12-16 15:42:53

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#14 2025-12-16 15:51:27

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#15 2025-12-16 16:02:25

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#16 2025-12-16 16:36:55

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#17 2025-12-16 17:02:24

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#18 2025-12-16 17:19:42

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#19 2025-12-16 17:24:05

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#20 2025-12-16 17:46:39

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#21 2025-12-16 18:05:59

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#22 2025-12-16 18:11:59

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#23 2025-12-16 18:39:24

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#24 2025-12-16 18:51:39

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

#25 2025-12-16 19:26:01

Re: [SOLVED] Soundcard not detected / Audio not working, new kernel.

Board footer