ASUS Vivobook S16 M3607GA — No sound card created (ACP70 + rt721-sdca, SoundWire enumerates but no ASoC machine driver match)

[mcvibu]

Hardware

• Laptop: ASUS Vivobook S16 M3607GA
• dmidecode system-product-name: Vivobook S16 M3607GA
• dmidecode baseboard-product-name: M3607GA
• Audio coprocessor: AMD ACP 7.0 (PCI 0000:03:00.5, [1022:15e2])
• Codec: Realtek RT721-SDCA over SoundWire

Kernel
6.19.10-300.fc44.x86_64 (Fedora 44)

Symptom
No internal speaker or headphone jack playback device is created.
aplay -l shows only HDMI outputs. Bluetooth audio works fine,
confirming the rest of the audio stack (PipeWire/WirePlumber) is healthy.

What I've confirmed

• SoundWire bus enumerates the codec correctly:
  /sys/bus/soundwire/devices/ shows sdw:0:1:025d:0721:01
  (Realtek rt721, matches SDCA functions UAJ/SmartMic/SmartAmp
  reported in dmesg at ACPI level)
• snd_pci_ps driver is bound to the PCI device
  (/sys/bus/pci/devices/0000:03:00.5/driver -> snd_pci_ps)
• SSDT10 ACPI table's acp-audio-config-flag is already 0x00 (Zero),
  so this doesn't look like the ASUS BIOS flag-override issue fixed
  for the Zenbook S16 UM5606GA.
• No ASoC machine driver ("acp_asoc_acp70.0" or similar) ever appears
  in dmesg, even with snd_sof.sof_debug=1 and
  snd_sof_amd.debug_mask=0xffff set. Reloading snd_pci_ps via modprobe
  produces no new log output either.
• All required modules (snd_sof_amd_acp70, soundwire_amd,
  snd_soc_rt721_sdca, snd_soc_sdca) are loaded. alsa-sof-firmware
  2025.12.2 is installed.

Suspected cause
Given the pattern in similar recent reports (ASUS TUF A14 FA401EA,
ASUS ProArt PX13 HN7306EA), this looks like a missing DMI quirk entry
for the M3607GA board in the ACP70 SoundWire machine-driver matching
path (possibly soc_sdw_quirk_table[] in
sound/soc/amd/acp/acp-sdw-legacy-mach.c, or an equivalent table),
similar to the PX13/A14 quirks merged in kernel 7.0. Since our SSDT
flag is already correct, I don't believe this is the same
acp-audio-config-flag override issue as the Zenbook S16 case.

Happy to test patches, run further debug commands, or provide any
additional logs needed.

[mcvibu]

Update: partial progress on kernel 7.1.5-201.fc44.x86_64 (Fedora 44)

Since filing this against 6.19.10, I've updated to kernel 7.1.5 and
the situation has changed — a sound card and machine driver binding
now appear where previously there was none.

aplay -l now shows:
card 1: amdsoundwire [amd-soundwire], device 0:
SDW1-PIN0-PLAYBACK-SimpleJack rt721-sdca-aif1-0
card 1: amdsoundwire [amd-soundwire], device 2:
SDW1-PIN1-PLAYBACK-SmartAmp rt721-sdca-aif2-2

wpctl status now shows real named sinks (not absent, not "Dummy
Output"):
Audio Coprocessor Speaker [vol: 0.37]
Audio Coprocessor Headphones [vol: 1.00]

However, internal speaker/headphone audio is still not audible.
amixer -c1 contents confirms nothing is muted and levels are at
100%/0dB (Speaker Switch: on, rt721 FU05/FU06 Playback Volume:
87/87). Testing both playback subdevices directly with aplay -D hw:1,0 and aplay -D hw:1,2 accepts stereo WAV playback
(VU meters move) but produces no audible output on either.

dmesg shows no firmware load failures for the SDCA functions
(SmartAmp/SmartMic/UAJ all enumerate cleanly at ACPI level), so
this doesn't look like the PX13-style missing-calibration-blob
case (no "firmware load ... failed with error -2" present).

So current state: machine driver now matches/binds (unlike 6.19),
mixer path is fully unmuted and accepting PCM, but the SmartAmp
output stage itself appears to never actually enable — possibly
still needs the board-specific amp-enable sequence even though
generic SDCA plumbing now works. Happy to test debug patches or
pull further SOF firmware logs if useful.