[DjDeveloperr]

Sound on the MacBook Pro 14" (M5 Pro, J714s) through the AOP, as on the J813: the T6050 AOP runs
the same audio service as T8142.

Rebased onto integration/aurora 58784a9: #1423, #1521 and #1557 (which carries the T6050 entry in
the T8152 AOP audio driver) are merged, so this PR is now two commits. The DT commit uses #1557's
audio node block (identity and DMA list) and adds the pins it needs to #1570's.

• ASoC: apple: macaudio: Add J714: the J813 configuration with two woofers and a tweeter per
  side. The speakers are behind the protection interfaces from the bind; each amp's level is its
  apple,speaker-config from the boot loader. KUnit tables and the binding updated.
• arm64: dts: apple: t6050-j714s: Add the speakers and the headphone jack: six SN012776 on
  i2c1/i2c3 (shared shutdown line and fault interrupt, SDZ fail-closed group, ti,bclk-ratio 250
  as on J813), the CS42L84 on i2c2, and the card. Slots, sense slots, idle masks and
  pins from the J714s ADT; only the pins these nodes use are exposed (pin 112 still SErrors).

Validated on the J714s:

• Lab boots of 7.3.0-m5-argb2 and, on the current default, 7.3.0-m5-intg2 (integration/aurora
  69a9fbb) with these three changes plus J813: AOP sound on integration/aurora (speakers, jack, microphones) #1423 and soc: apple: aop: Boot the T6050 (M5 Pro/Max) AOP through its setup port #1521 as modules. The intg2 runs booted this
  branch's own t6050-j714s.dts (plus the unit's GPU leakage words, as the default DTB does) on aura
  Stage 2 ca05ee02, which passes each amp its ADT speaker-config and amp-tx-zd-config row.
• Card AppleJ714 with the jack, speaker and sense PCMs; the six amps and the CS42L84 probe;
  nothing deferred. Each amp's playback slot (0-5) and V-sense slot read back as the ADT gives
  them; at bind all six are in software shutdown with no latched fault, the speaker volumes are held
  at the lock's ceiling and every amp gain at its ADT level until a protection owner takes them.
• speakerguardd v2 with this Mac's own tuning: its silent switch test passes with the wiring
  derived from this DT (on both kernels), v2 takes the speakers under the lease, and the supervisor
  grants only while a stream plays.
• Audible, owner present: a short tone at 5% and then 10% of the volume slider, each only after
  the safety gates passed (v2 owning the speakers and ready, amp temperatures and fault latches,
  nothing else playing). The kernel granted protection for the playback, the speaker and sense PCMs
  ran, a USB microphone picked up both tones above the quiet baseline, and the amps stayed at
  32-36 °C with no fault.

After the rebase (11-11): the rebased DT's audio nodes (card, amps, jack codec, regulator, buses, pins)
compare identical to the validated DTB's, phandles resolved, apart from #1557's hpai channel and
#1570's pin 81. A silent check on hardware with the rebased DT (intg2 lab boot; #1570's port-1 nodes
disabled there, since intg2 lacks its shared-PHY fix) brought up the speaker card and #1557's
microphone card and passed speakerguardd v2's dry-run check (switch test and verify).

Headphone jack (tested after the merge, same intg2 lab boot): insertion detected (jack switch, input
device, codec status); a 440 Hz tone at -20 dBFS with the jack DAC at -30 dB played clean (no xruns, no
codec errors) and a USB microphone heard it; the speakers stayed off throughout. Headset-mic capture
untested (the headphones have no mic).

🤖 Generated with Claude Code

[DjDeveloperr]

Owner-confirmed audible playback on the J714s: with the machine's owner at the M5, a short tone (440 Hz, then 1 kHz, 4 s) played through speakerguardd v2 at 5% and then 10% of the volume slider, and the owner confirmed hearing both tones, clean, with no distortion.

Setup: the current default kernel 7.3.0-m5-intg2 (integration/aurora 69a9fbb) with this PR's three changes plus #1423 and #1521 as modules, booting this branch's own t6050-j714s.dts (plus the unit's GPU leakage words, as the default DTB does) on aura Stage 2 ca05ee02. speakerguardd v2 ran with this Mac's own tuning (with speakerguardd#158's transport entry).

• Each step only after every gate passed: v2's silent switch test and verify, v2 owning the speakers and its supervisor ready, all six amps' temperature and fault latches, no other output stream.
• The kernel granted speaker protection for the playback; the speaker and sense PCMs ran; a USB microphone picked up both tones above the quiet baseline.
• Amp temperatures (°C), before / after 5% / after 10%: left woofer 1 32/32/32, left woofer 2 32/33/33, left tweeter 35/36/36, right woofer 1 33/33/33, right woofer 2 33/33/33, right tweeter 34/35/35. No fault latched.
• Ended at volume 0, muted.

Still not exercised: the headphone jack's streams (the CS42L84 probes and binds).

[DjDeveloperr]

Headphone jack, tested after the merge (11-11, lab boot of 7.3.0-m5-intg2 with this branch's DT, headphones without a mic plugged in):

• Detection: the card's Headphone Jack switch and the jack input device report the headphones, Headset Mic Jack stays off, and the CS42L84's plug status agrees.
• Playback: a 3 s 440 Hz tone at -20 dBFS through the jack PCM, with Jack DAC Playback Volume at -30 dB (read back before playing). The stream ran clean: no xruns and no codec or DMA errors in the log. A USB microphone near the headphones picked it up 35 dB above the quiet baseline in the 440 Hz band. The DAC went back to its minimum afterwards.
• Speakers stayed off for the whole test: no protection grant, speaker PCM closed, checked before and during playback.
• Capture: not tested, because these headphones have no mic.

So the jack's detection and playback path work on the J714s; headset-mic capture is still untested.

🤖 Generated with Claude Code