Skip to content

arm64: dts: apple: t6050: describe the MTP only on the laptops - #2507

Merged
DjDeveloperr merged 1 commit into
integration/aurorafrom
dtfill-t6050/mtp-default
Nov 20, 2026
Merged

DjDeveloperr merged 1 commit into
integration/aurorafrom
dtfill-t6050/mtp-default

Conversation

@RyanTheTide

Copy link
Copy Markdown
Contributor

Problem. t6050-max.dtsi describes the MTP coprocessor (keyboard and trackpad), its mailbox, its DART and its DockChannel FIFO with no status, so every board that includes the file has them enabled. The desktops don't have this block: the J775c and J873s ADTs have no mtp, dart-mtp or dockchannel-mtp. On J775c with CONFIG_APPLE_DOCKCHANNEL=y, dockchannel_probe() takes an SError and the kernel panics at about 0.14 s (reported and retested by Satya on #2495; the J775c-only fix is folded into #2102).

Change. The four nodes default to disabled in t6050-max.dtsi. Each board file that adds the HID transport enables them: t6050-j714c.dts, t6050-j714s.dts, t6050-j716c.dts, t6050-j716s.dts, and the shared t6050-k114-k116.dtsi (K114c, K114s, K116c, K116s). The MTP DART is part of the same block and is missing from the desktop ADTs too, so it follows the other three.

Testing. All 11 T6050 DTBs build with 0 dtc warnings before and after. Decompiled diff:

  • 8 laptops: only status = "okay" added on the four nodes. They had no status before, which already means available, and no phandle moves. So nothing changes on these boards.
  • J775c, J873s: the four nodes become disabled.
  • J775d: byte-identical (it doesn't include t6050-max.dtsi).

dt-validate with the tree's schema: the same 18 messages before and after, 0 new.

Still to run.

  • K114c: keyboard and trackpad still work at this head (dmesg | grep -i -E "mtp|dockchannel|hid", type and move the pointer). @joshuaswarren's agent offered this row.
  • J873s: boots past dockchannel_driver_init with CONFIG_APPLE_DOCKCHANNEL=y (no known owner yet).
  • J714s needs no row: Dj checked the same change on arm64: dts: apple: support M5 Max Mac Studio (J775c) #2102 and it produces no effective DTB change there.
Per-board DTB diff and how it was built

Base integration/aurora 2c35f50, commit 712f23d. Built with the kbuild cpp + dtc flags (scripts/Makefile.dtbs warning set), decompiled with dtc -I dtb -O dts -s.

Board Changed lines in decompiled DTB
J714c, J714s, J716c, J716s, K114c, K114s, K116c, K116s 4 x + status = "okay"; (mailbox@294608000, iommu@294800000, mtp@294600000, fifo@294b14000)
J775c, J873s 4 x + status = "disabled"; (same nodes)
J775d none

ADT check (each board's own ADT, node names only): laptops have mtp, dart-mtp and dockchannel-mtp; J775c, J775d and J873s have none of them.

#2102 carries the J775c-only fold as d554f54 (J775c there is a flattened board file). The two don't conflict: this change covers J873s and the integration J775c file, and #2102's J775c file sets the same status itself.

Assisted-by: LLM (Claude:Opus 5.5)

t6050-max.dtsi describes the MTP coprocessor (the keyboard and trackpad's),
its mailbox, its DART and the DockChannel FIFO its HID transport uses, with
no status, so every board that includes the file gets all four enabled.

Only the laptops have this block. The ADTs of J714c, J714s, J716c, J716s,
K114c, K114s, K116c and K116s carry mtp, dart-mtp and dockchannel-mtp; the
Mac Studio (J775c) and Mac mini (J873s) ADTs carry none of them, and
nothing brings the MTP up there. On a J775c built with
CONFIG_APPLE_DOCKCHANNEL=y the first register access in dockchannel_probe()
takes an SError and the kernel panics at about 0.14 s.

Mark the four nodes disabled in t6050-max.dtsi and enable them in each
board file that adds the HID transport: t6050-j714c.dts, t6050-j714s.dts,
t6050-j716c.dts, t6050-j716s.dts and the shared t6050-k114-k116.dtsi
(K114c, K114s, K116c, K116s). The MTP DART is included because it is part
of the same block and is absent from the desktop ADTs as well.

On the eight laptops the compiled DTBs differ only by an explicit
status = "okay" on the four nodes, which had no status before, so their
effective state is unchanged. J775c and J873s get the four nodes disabled.
J775d does not include t6050-max.dtsi and is unchanged.

Reported-by: Satya Benson <satya@benson.earth>
Suggested-by: DjDeveloperr <43033058+DjDeveloperr@users.noreply.github.com>
Suggested-by: Joshua Warren <816217+joshuaswarren@users.noreply.github.com>
Link: #2102
Link: #2495
Signed-off-by: Ryan Murray <ryan@aurorasilicon.org>

@DjDeveloperr DjDeveloperr left a comment

Copy link
Copy Markdown
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Approving. This is the hardware-description shape: the MTP, its mailbox, DART and DockChannel FIFO are off in t6050-max.dtsi and on in the boards that wire a keyboard and trackpad (J714c, J714s, J716c, J716s and the K114/K116 include). Including mtp_dart is right, since the desktop ADTs lack dart-mtp too. On the laptops the DTB only gains explicit okays: the J714s DTB diff from the M5 lane showed exactly that. J775c and J873s stop probing hardware they don't have, which is the DOCKCHANNEL=y SError Satya hit.

Merge once dt-checks is green and Joshua's K114c DTB before/after diff shows only the four added okays. The booted K114c keyboard and trackpad row and a J873s boot row can follow after merge.

@DjDeveloperr
DjDeveloperr merged commit 4c4cbf0 into integration/aurora Nov 20, 2026
2 checks passed
@DjDeveloperr

Copy link
Copy Markdown
Contributor

J714s (M5 Pro) DTB check at this head: no effective change, so no hardware row.

Claude Opus 5.5, from DjDeveloperr's M5 lane.

What was built. The J714s DTB at head 712f23deb5c2 and at its merge base on integration/aurora, 2c35f5094de1. Same method as the #2102 check: the kernel's cmd_dtc steps (gcc -E -nostdinc -undef -D__DTS__, then the in-tree DTC 1.7.2-g53373d13 with the scripts/Makefile.dtbs warning set), decompiled with dtc -I dtb -O dts, sorted (-s) and unsorted.

J714s DTB bytes sha256
merge base 2c35f5094de1 239,719 e87b7434d55fc1173bc7457cd55062860d2411e3e0a32a76fe70352109c1d659
this head 712f23deb5c2 239,799 7872c5a36895addd6da1ee8892874f1518fa128b80a69cd192716cb11c6304b5

The whole decompiled diff (sorted; the unsorted diff is the same four lines):

@@ fifo@294b14000 {            /* mtp_dockchannel */
 			reg-names = "irq";
+			status = "okay";
@@ iommu@294800000 {           /* mtp_dart */
 			reg = <0x02 0x94800000 0x00 0xc000>;
+			status = "okay";
@@ mailbox@294608000 {         /* mtp_mbox */
 			reg = <0x02 0x94608000 0x00 0x4000>;
+			status = "okay";
@@ mtp@294600000 {             /* mtp */
 			reg-names = "asc", "sram";
+			status = "okay";

The 80-byte size difference is exactly these four properties. No phandle is renumbered, and every reference into the block is byte-identical: mtp keeps mboxes = <&mtp_mbox> and iommus = <&mtp_dart 0>, and the HID child input@8000 keeps iommus = <&mtp_dart 0>, apple,helper-cpu = <&mtp> and interrupt-parent = <&mtp_dockchannel>. The MTP DART stays enabled and referenced on J714s.

Effective state is the same. The four nodes had no status before, which __of_device_is_available() treats as available, and now they say "okay". aura's status gates treat a missing status and "okay" the same way, including the DAPF gate in dapf.c that decides whether dart-mtp gets its DMA setup and the missing-device and reserved-memory passes in kboot.c.

Verdict. No J714s row needed for this head. This covers J714s only. The other laptops' opt-ins and the J873s boot row are listed above in the PR description.

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