Skip to content

arm64: dts: apple: t6050-j714c: Trackpad reset, SMC sensors and keyboard backlight - #1984

Open
XY55bqg6iAw8ttP wants to merge 3 commits into
aurora-silicon:integration/aurorafrom
XY55bqg6iAw8ttP:j714c-dt-fixes
Open

XY55bqg6iAw8ttP wants to merge 3 commits into
aurora-silicon:integration/aurorafrom
XY55bqg6iAw8ttP:j714c-dt-fixes

Conversation

@XY55bqg6iAw8ttP

Copy link
Copy Markdown

Three small fixes for the 14-inch M5 Max (J714c), all in t6050-j714c.dts. Each one copies what the J714s already does, with the values checked against a J714cAP ADT (macOS 27.3 beta 1, 26C5035r). The other ten T6050 board DTBs are byte-identical after each commit.

What changes

Commit Problem on J714c today Fix
Reset the trackpad AFE from the host apple,firmware-managed-reset leaves the AFE reset to the MTP firmware, which never does it on this board: AFE[0] Chip Boot Failure for HBPP: No HINT_L, then iface multi-touch start timed out. The trackpad is dead. apple,afe-reset-gpios = <&smc_gpio_low 28 GPIO_ACTIVE_LOW> on mtp_hid, as on the J714s, and drop apple,firmware-managed-reset.
Enable the SMC fan, power and temperature sensors &smc_hwmon is disabled ("fan and power keys are not yet checked"), so no fan, power or temperature readings. The J714s block: hwmon-common, hwmon-fan-dual, hwmon-laptop, fan mode keys F0md/F1md, temperature discovery.
Describe the keyboard backlight No kbd_backlight LED. pwm-leds on fpwm0 channel 0, 40000 ns, as on the J714s.

Where the values come from (J714cAP ADT)

  • Trackpad reset: dockchannel-mtp/mtp-transport/multi-touch function-afe-reset = [0x293 'pIcp' 0x10000 0x11c0000], the same SMC pcIO template as the J714s, so SMC GPIO gp1c (28), active low.
  • Fans: two fans, F0Ac/F1Ac, 2317 to 7826 rpm; the SMC drives them.
  • Keyboard backlight: arm-io/pwm0 is fpwm,s5l8920x at reg 0x5040000 0x3 (0x5_05040000 in the DT, the same arm-io offset as the GPU), pwm0/kbd-backlight default-hz = 25000, so a 40000 ns period.

Tested

On a J714c (Mac17,7, M5 Max), macOS 27.3 beta 1 firmware, kernel 7.3.0-k114c, with the DTB built from this branch:

  • Trackpad: New AFE[0] cbor image received, then Interface multi-touch is now ready, and the pointer moves.
  • macsmc_hwmon: two fans, 342 discovered temperature keys, Total System Power, AC input and heatpipe power.
  • /sys/class/leds/kbd_backlight: max 255; at 128 the PWM runs 20 us / 40 us and the keys light. systemd-backlight restores the level at boot.

The same three changes also ran for a day as an overlay on Warren's k114c-m5max-kernel J714c DT.

🤖 Generated with Claude Code

XY55bqg6iAw8ttP and others added 3 commits November 15, 2026 12:19
With apple,firmware-managed-reset the MTP firmware is left to reset the
trackpad's AFE, and on the J714c it never does: the AFE never boots
("AFE[0] Chip Boot Failure for HBPP: No HINT_L") and the multi-touch
interface times out, so the trackpad is dead.

The J714cAP ADT gives dockchannel-mtp/mtp-transport/multi-touch a
function-afe-reset of the SMC pcIO template (0x10000, 0x11c0000), the
same line as on the J714s: SMC GPIO gp1c, active low. Describe it as
apple,afe-reset-gpios on mtp_hid, as the J714s does, and drop
apple,firmware-managed-reset.

Tested on a J714c: the host sends the AFE image, the multi-touch
interface comes up and the trackpad works. The other ten T6050 board
DTBs are byte-identical.

Assisted-by: LLM
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: XY55bqg6iAw8ttP <132220364+XY55bqg6iAw8ttP@users.noreply.github.com>
…ture sensors

The J714c left smc_hwmon disabled until its fan and power keys were
checked. The J714cAP SMC has the J714s's keys: two fans, F0Ac and F1Ac,
2317 to 7826 rpm, with the T6050 mode keys F0md and F1md, and the shared
power, voltage and current keys. Use the J714s block: the common, dual
fan and laptop sensors, temperature discovery, and the shared
temperature nodes and apple,soc-die-keys removed so discovery runs.

Tested on a J714c: macsmc_hwmon reports both fans, 342 temperature keys
and the system, AC input and heatpipe power. The other ten T6050 board
DTBs are byte-identical.

Assisted-by: LLM
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: XY55bqg6iAw8ttP <132220364+XY55bqg6iAw8ttP@users.noreply.github.com>
The J714cAP ADT has arm-io/pwm0, an fpwm s5l8920x at 0x3_05040000
(0x5_05040000 here, fpwm0 in the T6050 inventory), with a kbd-backlight
child at default-hz 25000. Describe it as the J714s does: a pwm-leds
kbd_backlight on fpwm0 channel 0 with a 40000 ns period, and enable
fpwm0.

Tested on a J714c: kbd_backlight appears with max brightness 255, the
PWM follows the brightness (20 us of 40 us at 128) and the keys light.
The other ten T6050 board DTBs are byte-identical.

Assisted-by: LLM
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Signed-off-by: XY55bqg6iAw8ttP <132220364+XY55bqg6iAw8ttP@users.noreply.github.com>
@RyanTheTide

Copy link
Copy Markdown
Contributor

@XY55bqg6iAw8ttP, thanks for the J714c work here. Beyond this PR, two J714c rows would help:

  1. Speakers: the J714c has no sound card node yet. Please post the ADT values for the amp nodes (i2c addresses, speaker order) so we can add it. The speakers stay at the safe level, so please don't play anything through them. Confirming that the amps probe is enough.
  2. SIO and the PMP CA-event ack: today both are enabled only on the J714s (drivers/dma/apple-sio.c:703-706, drivers/soc/apple/rtkit.c:2276-2277). We are widening both to every board with the hardware and will link the branch. A boot of it plus dmesg | grep -i -E "rtkit|pmp|sio" before and after 30 minutes of uptime is what we need: SIO users (DP audio) probe, and no RTKit timeout.

dmesg plus the command output, here or in a new issue, is perfect.

@RyanTheTide

Copy link
Copy Markdown
Contributor

As promised, the branch that widens SIO and the PMP CoreAnalytics acknowledgement from the J714s to every T6050 board (keyed on the hardware instead of the board name) is #2412 (draft). A boot of it plus dmesg | grep -i -E "sio|rtkit|pmp" right after boot and after 30 minutes of uptime is all we need: SIO users probe, no "refused auxiliary buffer request", no RTKit timeouts.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants