Update: found a reproducible trigger, captured the failed state, and re-
verified the cure.

Trigger: a warm reboot round-trip through Windows (dual-boot machine:
warm-restart to Windows, use the microphone there, warm-restart back to
Ubuntu) brings the microphone up in the corrupted state — full-scale
impulse train, 3953 large sample-to-sample discontinuities in the
attached ~5.8 s capture. Reproduced deliberately on the first attempt;
will re-verify repeatability.

Cure verified in the same cycle: clean state confirmed → Windows round-
trip → corruption confirmed and logged → cold power-off → clean capture
confirmed. This matches all earlier persistence findings (the corrupted
state survives warm reboots, suspend/resume, and a full
SOF/SoundWire/codec module reload; only cold power-off clears it).

Interpretation: Windows' audio driver leaves the SoundWire codec/link in
a state that the Linux stack does not fully re-initialize on warm boot.
Since this machine is Ubuntu-certified and dual-boot is a common
configuration, this seems squarely in scope.

Attaching: alsa-info captured in the failed state, full dmesg of the
affected boot (no sof/soundwire errors logged — the corruption is
silent), and the corrupted capture sample.


** Attachment added: "Archive.zip"
   
https://bugs.launchpad.net/ubuntu/+source/linux-hwe-6.14/+bug/2160386/+attachment/5981173/+files/Archive.zip

-- 
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2160386

Title:
  Dell Precision 5690 (Ubuntu-certified, MTL): internal DMIC capture
  corrupts to full-scale impulse train; state survives warm reboot,
  cleared only by cold power-off

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux-hwe-6.14/+bug/2160386/+subscriptions


-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to