Skip to content

arm64: dts: apple: t8142-j815: MagSafe 3 PD controller - #1907

Merged
RyanTheTide merged 2 commits into
aurora-silicon:integration/aurorafrom
m0d16l14n1:j815/t8142-j815-magsafe
Nov 22, 2026
Merged

RyanTheTide merged 2 commits into
aurora-silicon:integration/aurorafrom
m0d16l14n1:j815/t8142-j815-magsafe

Conversation

@m0d16l14n1

@m0d16l14n1 m0d16l14n1 commented Nov 14, 2026 •

Copy link
Copy Markdown
Contributor

Adds the J815's MagSafe 3 PD controller, as #1686 did for the J714s.

From the J815 ADT: hpm5 on nub-spmi-a0, SID 8, port-type 0x11 (MagSafe), features-supported <3> ("Power In" only).

  • No USB data path, so no connector node: tipd registers a Type-C port and a power supply for it.
  • The T8142 SPMI bus is polled, so like hpm0 and hpm1 it has no interrupts.
  • apple,no-liquid-detection (usb: typec: tipd: sn201202x: skip liquid detection where the port has none #1685): the port has no LDCM.
  • apple,sleep-ssps-rules, as hpm0 and hpm1 now have in integration/aurora (6465f0a) and as the J714s's MagSafe hpm5 has. So all three J815 controllers are talked to under the sleep's command rules. It doesn't get apple,sleep-ssps: no SSPS run has been made on it, and a controller that refuses S3 refuses the sleep.

Changed on this head (review): rebased onto integration/aurora cc14570 (the conflict was next to the new hpm0/hpm1 rules), and apple,sleep-ssps-rules added to hpm5.

Tested on J815 (MacBook Air 15" M5), macOS 27.2 stub, on the earlier head:

  • sn201202x binds SPMI 1-08. The J815 now has 3 Type-C ports (hpm0 1-0c, hpm1 1-0a, hpm5 1-08).
  • New power supply tps6598x-source-psy-1-08.

Checks on this head: the J813/J815 DTBs build, dt_binding_check and CHECK_DTBS with apple,sn201202x.yaml are clean, and checkpatch --strict is clean.

Still owed (HW row): J815 on the rebased kernel and DTB, with the default command line. With a MagSafe charger attached, tps6598x-source-psy-1-08 reports online=1 with a sane voltage and current, and an unplug and replug updates it. Then one s2idle cycle with the charger and one without, with no tipd errors on resume; 0 SELinux denials, 0 failed units, 0 new kernel errors.

Part of the J815 bring-up: board results are summarised in discussion #135.

Assisted-by: LLM (Claude:Opus 5.5)

@m0d16l14n1

Copy link
Copy Markdown
Contributor Author

@DjDeveloperr out of draft, parent-first on #1532 (alias fix merged in). On the 27.2 stub: sn201202x at 1-08, all three Type-C ports, and the MagSafe power supply registered. Please review.

Assisted-by: Claude Opus 5.5

@m0d16l14n1
m0d16l14n1 force-pushed the j815/t8142-j815-magsafe branch from 7a196b1 to c3e48b3 Compare November 19, 2026 07:27
@m0d16l14n1 m0d16l14n1 changed the title arm64: dts: apple: t8142-j815: MagSafe 3 PD controller (on #1532) arm64: dts: apple: t8142-j815: MagSafe 3 PD controller Nov 19, 2026
@RyanTheTide

Copy link
Copy Markdown
Contributor

Blockers: 2.

The node matches the J714s MagSafe controller (same compatible, SID, no interrupts on the polled bus, apple,no-liquid-detection), and the ADT facts in the comment support it.

  1. The branch conflicts with integration/aurora. Since "usb: typec: tipd: the sleep's SSPS rules and the probe reset from the DT" (6465f0a), the sleep's system power state rules are no longer chosen by a board list that included the J815. Each controller now needs apple,sleep-ssps-rules. integration/aurora adds it to &hpm0 and &hpm1 at the end of arch/arm64/boot/dts/apple/t8142-j815.dts, and the J714s gives its MagSafe hpm5 the property too. After the rebase, hpm5 here (t8142-j815.dts:483-487) would be the only J815 controller without it, so it would handle sleep differently from the other two ports. Add apple,sleep-ssps-rules; to hpm5.
  2. The commit carries an Assisted-by: Claude Opus 5.5 trailer. Commits should carry only the human Signed-off-by. Put the disclosure as the PR body's last line in the form Assisted-by: LLM (Claude:Opus 5.5).

HW row: J815 on the rebased kernel and DTB, default command line: with a MagSafe charger attached, tps6598x-source-psy-1-08 reports online=1 with a sane voltage/current, and unplugging and replugging updates it. One s2idle cycle with the charger attached and one without, with no tipd errors on resume. 0 SELinux denials, 0 failed units, 0 new kernel errors.

Assisted-by: LLM (Claude:Opus 5.5)

hpm5 on nub-spmi-a0 (SID 8) is the MagSafe 3 port's PD controller (ADT
port-type 0x11). It has no USB data path, so it gets no connector node;
tipd registers a Type-C port and a power supply for it, as on J714s. The
bus is polled on T8142, so like hpm0 and hpm1 it has no interrupts. Its
ADT features-supported is <3> ("Power In" only), so it is marked
apple,no-liquid-detection.

Like hpm0 and hpm1, it takes the sleep's command rules
(apple,sleep-ssps-rules), as the J714s's MagSafe controller does. It does
not get apple,sleep-ssps: no SSPS run on it has been made yet, and a
controller that refuses S3 refuses the sleep.

Signed-off-by: Nick <nekreshen@yandex.ru>
@m0d16l14n1
m0d16l14n1 force-pushed the j815/t8142-j815-magsafe branch from c3e48b3 to 6dd5889 Compare November 21, 2026 12:51
@m0d16l14n1

Copy link
Copy Markdown
Contributor Author

@RyanTheTide fixed.

  • Rebased onto cc14570.
  • hpm5 now has apple,sleep-ssps-rules, next to its other properties.
  • The commit has only my sign-off.

hpm5 doesn't get apple,sleep-ssps: no SSPS run has been made on it, and a refused S3 refuses the sleep. HW row (MagSafe online, replug, s2idle with and without the charger) to follow.

Assisted-by: LLM (Claude:Opus 5.5)

@RyanTheTide

Copy link
Copy Markdown
Contributor

Re-review of 6dd5889: No blockers.

Earlier points: 1 resolved: hpm5 now has apple,sleep-ssps-rules (arch/arm64/boot/dts/apple/t8142-j815.dts:545), like &hpm0 and &hpm1 just above it and the J714s MagSafe controller, and the branch is rebased onto the integration branch. Leaving out apple,sleep-ssps until an SSPS run has been made on this controller is the safe choice. 2 resolved: the commit carries only the human Signed-off-by.

HW row: J815 on this head, default command line: with a MagSafe charger attached, tps6598x-source-psy-1-08 reports online=1 with a sane voltage/current, and unplugging and replugging updates it. One s2idle cycle with the charger attached and one without, with no tipd errors on resume. 0 SELinux denials, 0 failed units, 0 new kernel errors.

Assisted-by: LLM (Claude:Opus 5.5)

@RyanTheTide

Copy link
Copy Markdown
Contributor

Merge-train note: #1806 landed in batch 1 (integration/aurora 606d37e) and conflicts with this branch in arch/arm64/boot/dts/apple/t8142-j815.dts. Review is still no blockers. Please merge integration/aurora into the branch (or rebase onto a new one); it goes into the next batch once clean and green.

Signed-off-by: Nick <nekreshen@yandex.ru>

# Conflicts:
#	arch/arm64/boot/dts/apple/t8142-j815.dts

@RyanTheTide RyanTheTide 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.

Re-checked at 8d16bfb. The merge keeps both sides in t8142-j815.dts: the MagSafe PD controller node, then the t8142-j813-j815-sleep.dtsi and t8142-j813-dcpext.dtsi includes from integration/aurora. Against the current tip the branch adds only its one file (+17), the commit reviewed before. Approving at this head.

Assisted-by: LLM (Claude:Opus 5.5)

@m0d16l14n1

Copy link
Copy Markdown
Contributor Author

@RyanTheTide merged integration/aurora (b241e35) into the branch, no force push: 8d16bfb3.

In t8142-j815.dts the conflict was the hpm5 node against #1806's two includes. hpm5 sits after &hpm1, followed by t8142-j813-j815-sleep.dtsi and t8142-j813-dcpext.dtsi. Against integration, the merge adds only the hpm5 block.

CHECK_DTBS on the J704/J813/J815 DTBs gives the same 4 warnings as integration (the existing disp0_piodma alias), none new.

Assisted-by: LLM (Claude:Opus 5.5)

@RyanTheTide
RyanTheTide merged commit 85e255d into aurora-silicon:integration/aurora Nov 22, 2026
2 checks passed
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