Re: Bug#1100928: [regression 6.1.y] microphone no longer records (on VivoBook_ASUSLaptop TP401MARB_J401MA)

[Salvatore Bonaccorso <carnil@debian.org>]

Hi Craig, all

On Sat, Mar 22, 2025 at 08:55:17AM +0100, Salvatore Bonaccorso wrote:
> Control: tags -1 + moreinfo upstream
> 
> Hi
> 
> On Thu, Mar 20, 2025 at 10:55:32AM -0400, C.D. MacEachern wrote:
> > Package: src:linux
> > Version: 6.1.129-1
> > Severity: important
> > X-Debbugs-Cc: craig.daniel.maceachern@gmail.com
> > 
> > Dear Maintainer,
> > 
> > After update to kernel image 6.1.0-32 on Debian bookworm my builtin microphone
> > would no longer record
> > anything - no levels detected and I checked for muted channels with tools like
> > pavucontrol and alsamixer.
> > 
> > I found that my microphone was using the `snd_hda_intel` driver in the kernel,
> > so tried rebooting and choosing
> > the 6.1.0-31 kernel instead to rule out hardware issue. Previous kernel works
> > as expected, mic records and playback
> > is fine, so some update related to this driver, or the driver itself was
> > updated and no longer works correctly with
> > the builtin microphone.
> 
> Thanks for reproting the issue (leaving boot log context, hw used,
> below for context).
> 
> I think this might be introduced with 3b4309546b48 ("ALSA: hda: Fix
> headset detection failure due to unstable sort") wich landed as well
> in 6.1.129.
> 
> If you revert that commit on top of 6.1.129, does that fixes your
> issue? Would you be able to test this?
> 
> #regzbot introduced: v6.1.128..v6.1.129
> #regzbot link: https://bugs.debian.org/1100928
> 
> The solution might be similar to c6557ccf8094 ("ALSA: hda/realtek: Fix
> microphone regression on ASUS N705UD") from 6.14-rc5 (which got
> backported to 6.13.6, 6.12.18, 6.6.81 but not yet 6.1.y).

In case you need help in trying a kernel build with the revet applied,
I'm attaching the revert patch. With that you can follow
https://kernel-team.pages.debian.net/kernel-handbook/ch-common-tasks.html#id-1.6.6.4
to build a kernel using the `test-patches` helper script to test with.

Is that enough help so we can confirm the breaking commit?

Kuan-Wei Chiu, Takashi Iwai any input on our issue?

Regards,
Salvatore