Skip to content

arm64: dts: apple: t6050: share the keyboard backlight, add it to J716c - #2282

Draft
RyanTheTide wants to merge 1 commit into
integration/aurorafrom
dtfill-t6050/j716c-kbdlight
Draft

RyanTheTide wants to merge 1 commit into
integration/aurorafrom
dtfill-t6050/j716c-kbdlight

Conversation

@RyanTheTide

@RyanTheTide RyanTheTide commented Nov 19, 2026 •

Copy link
Copy Markdown
Contributor

Problem

The J716c has no keyboard backlight. t6050-j716c.dts leaves fpwm0 disabled and has no pwm-leds node. Its ADT has the same pwm0 kbd-backlight node as the J714s, the J716s and the K114/K116 boards: channel 0 of fpwm0 at 25 kHz.

The J714s and J716s carry identical copies of the LED block. The K boards have their own copy, with a different label.

Change

  • New t6050-kbd-backlight.dtsi: the pwm-leds node (pwms = <&fpwm0 0 40000>, label kbd_backlight, default-state = "keep") and &fpwm0 { status = "okay"; }.
  • t6050-j714s.dts, t6050-j716s.dts: the copy is replaced by the include.
  • t6050-k114-k116.dtsi: includes the file. A small override keeps the existing apple::kbd_backlight label and drops default-state, so the K boards' DTBs do not change. The rear logo LED stays in the same controller.
  • t6050-j716c.dts: includes the file. This is the only functional change.

J714c: #1984 adds the same LED block to t6050-j714c.dts, with the same label, default-state and period. It can rebase onto this PR and use #include "t6050-kbd-backlight.dtsi" instead. This PR does not touch J714c.

Testing

  • Full kernel tree make dtbs CHECK_DTBS=y (dtschema) for every T6050 DTB on this branch: builds clean, no dtc warnings, and no schema message that integration/aurora does not already have.
  • All 11 T6050 DTBs build with no dtc warnings, before and after.
  • J714c, J714s, J716s, J775c, J775d, J873s, K114c, K114s, K116c and K116s DTBs are byte-identical.
  • J716c gains the led-controller node, and its fpwm0 goes to okay (diff below). ps_fpwm0 is already on.
  • dt-validate on J716c: no new messages (14 before, 14 after, all pre-existing opp-table and pcie messages).

Still to run

  • J716c (erikwb): brightnessctl -d kbd_backlight set 50% lights the keys. The level is the same after an s2idle cycle.

No loader change is needed. It does not depend on #2208/#2209. Those touch other nodes in t6050-j716c.dts, so a rebase there is trivial.

Per-board values and DTB diffs

Values come from each board's ADT: pwm0 kbd-backlight, channel 0, 25 kHz, so a 40000 ns period. This is the same on J714c, J714s, J716c, J716s and K116s (checked in each board's ADT) and on the other K boards (already described that way).

J716c, decompiled DTB, phandle renumbering masked:

+	led-controller {
+		compatible = "pwm-leds";
+		led-0 {
+			color = <0x00>;          /* LED_COLOR_ID_WHITE */
+			default-state = "keep";
+			function = "kbd_backlight";
+			label = "kbd_backlight";
+			max-brightness = <0xff>;
+			pwms = <&fpwm0 0x00 0x9c40>;
+		};
+	};
 ...
 pwm@505040000 {
+	phandle = <...>;
-	status = "disabled";
+	status = "okay";
 };

All other boards: byte-identical DTB.

Assisted-by: LLM (Claude:Opus 5.5)

The J716c leaves fpwm0 disabled and has no keyboard backlight LED. Its
ADT has the same pwm0 kbd-backlight node as the J714s, J716s and the
K114/K116 boards: channel 0 of fpwm0 at 25 kHz.

Move the pwm-leds node and the fpwm0 enable, which the J714s and J716s
each carried as an identical copy, into a new t6050-kbd-backlight.dtsi.
Include it from the J714s, J716s, J716c and t6050-k114-k116.dtsi. The
K boards keep their own LED label and no default-state through a small
override, and keep their rear logo LED in the same controller.

The J716c gains the LED and an enabled fpwm0. The J714s, J716s, K114c,
K114s, K116c and K116s DTBs are byte-identical; no other T6050 board
changes. The J714c can take the same include.

The J714s block was added by DjDeveloperr, the J716s one by Rolando
Trevino.

Signed-off-by: Ryan Murray <ryan@aurorasilicon.org>
@RyanTheTide

Copy link
Copy Markdown
Contributor Author

No blockers.

Rebuilt every T6050 DTB on current integration/aurora: only J716c changes, gaining the pwm-leds node and fpwm0 okay (ps_fpwm0 is already on). The other ten are byte-identical, including the K boards, where the override keeps apple::kbd_backlight and drops default-state. dtc warning counts are unchanged and dt-checks pass. host-tests failed on the apt mirror; rerun it. The node uses the same fpwm0 channel and 40000 ns period as J714s/J716s, from the same ADT pwm0/kbd-backlight entry. Can leave draft.

  1. Nit: t6050-k114-k116.dtsi keeps a second name (apple::kbd_backlight, no default-state) for the same LED. UPower matches both names, but a follow-up that settles the T6050 laptops on one name and one boot state would remove the override.

HW row: J716c on the head: brightnessctl -d kbd_backlight set 50% lights the keys, the level holds across one s2idle, 0 AVC, 0 failed units, no new kernel errors.

Assisted-by: LLM (Claude:Opus 5.5)

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.

1 participant