Skip to content

Offer Touch ID on the 16-inch M4 Max (J616c) - #160

Open
allanclempe wants to merge 4 commits into
omacom:mainfrom
allanclempe:oms-2-touchid-j616c
Open

allanclempe wants to merge 4 commits into
omacom:mainfrom
allanclempe:oms-2-touchid-j616c

Conversation

@allanclempe

@allanclempe allanclempe commented Nov 18, 2026 •

Copy link
Copy Markdown

Note

Depends on #140 (head b4d5c2f). The first commit is #140; the last three are this change. Merge #140 first; this PR then shows only its own commits.

Summary

Extends #140's Touch ID for the J616s to the 16-inch M4 Max (J616c).

The J616c has the J616s's sensor, wiring and enclave profile on its own SoC: aurora's SEP driver runs it under T6041/J614c, J616c, which is T6040's profile. Its sensor sits at the root of the tree (/spi@429108000/fingerprint@0), and its node names apple/mesacal-j616c.bin and apple/msrk-j616c.bin, both sealed for chip 00006041. Without this, touchid-eligibility prints none on the J616c.

 touchid-eligibility
+  one sealed_board check: apple,j616s on T6040, or apple,j616c on T6041
+  the 14-inch J614c stays unsupported until tried
 touchid-enable
+  J616c: writes xart_guard=1 with the driver's options (also enable-fingerprint)
+  J616c: reseals mesacal-j616c.bin (FSC2) and msrk-j616c.bin (MSRk) with --chip 00006041
 touchid-root-key
+  accepts --chip 00006041
 disable-fingerprint
+  deletes msrk-j616c.bin too

Evidence

On a J616c (7.3.0.aurora1-40, omarchy-mac 0.2.3-2 with these four scripts, libfprint-git from omacom/omarchy-pkgs#2737's aarch64 CI build), through an unchanged omarchy-setup-security-fingerprint:

  • Eligibility: touchid-eligibility prints available (before: none).
  • Setup: consent; both files rebuilt from the factory records on the ESP, root-only; then enrol and ✓ Matched.
  • Use: sudo and the lock screen with a finger. The setup configured polkit too; a polkit prompt hasn't been tried yet.
  • After a restart: the driver logs "restored 1 enrolled identity/identities; enrolment survived reboot", and sudo and the lock screen match again.
  • The guard: the trial ran with options apple_sep xart_guard=1. xart_guard_refusals stayed 0 through keybag creation, enrolment and matching. At boot the driver lists 10 session records, 0 created by Linux, and protects all 10. The option was set by hand for the trial. 7.3.0.aurora1-40 runs the J616c with the guard off (diag/xart_guard 0), although T6041's profile in aurora-silicon/linux's current source has guard 1, so this PR writes xart_guard=1 on a J616c beside the other two options, in touchid-enable and in enable-fingerprint. Other Macs keep the kernel's level. A later kernel with guard 1 makes it a no-op.

Tests: test/touchid-consent-test.sh gains a J616c section:

  • eligibility, including without the factory records;
  • both files rebuilt root-only;
  • records sealed for the M4 Pro refused;
  • a J614c, and a J616s on T6041, unsupported;
  • disable deleting the key.

The section fails when the J616c is given the M4 Pro's chip. omarchy-mac/test/all passes, 124 ok. shellcheck (-S warning) reports nothing new against #140's head.

Merge Danger

Door: two-way

Merges with #140, after it. Touch ID on the J616c also needs a libfprint-git that accepts SEP interface v5 (#2737 on omarchy-pkgs); without it the setup can't enrol, as on any Mac on the 7.3 kernel.

Blast Radius: J616c

Other Macs keep their behaviour; the J616s takes the same paths as in #140.

scottjones and others added 2 commits November 14, 2026 11:30
The J616s read "none": its sensor sits on an SPI controller at the root
of the device tree, not under /soc, and its SoC wasn't on the list.
Like the Neo, its sensor loads a sealed root key, and its calibration is
sealed too. Its node names both (mesacal-j616s.bin, then
msrk-j616s.bin), and the calibration name was read as both entries run
together.

- touchid-eligibility finds the sensor at the root as well, admits the
  J616s, and reads each firmware-name entry on its own.
- touchid-root-key takes the SoC the seal must name (--chip) and the
  record kind (--record MSRk or FSC2). The default stays the Neo's key.
- touchid-enable rebuilds the J616s calibration and root key from its
  factory records on the ESP, checked against its iBoot System
  Container, and refuses a firmware name outside apple/.
- disable-fingerprint deletes either sensor key.

Tested:
- On a J616s, touchid-root-key rebuilds both files byte for byte as
  aurora-silicon's aurora-touchid-setup does. With those files Touch ID
  enrols and verifies, also after a cold boot.
- The consent test gains a J616s section, which fails on the old scripts.

Assisted-by: Claude:claude-opus-5-5
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The J616c carries the J616s's sensor, wiring and enclave profile on its own
SoC: aurora's SEP driver runs it under T6041, which is T6040's profile. Its
sensor sits at the root of the tree and its node names
apple/mesacal-j616c.bin and apple/msrk-j616c.bin, both sealed for 00006041.

- touchid-eligibility admits apple,j616c on apple,t6041 beside the J616s,
  through one sealed_board check; the 14-inch J614c stays unsupported until
  tried.
- touchid-enable rebuilds both files with --chip 00006041.
- touchid-root-key accepts 00006041.
- disable-fingerprint deletes msrk-j616c.bin too.

Tests: a J616c section in touchid-consent-test.sh (eligibility, both files
rebuilt root-only, an M4 Pro seal refused, a J614c and a J616s on T6041
unsupported, disable deleting the key). It fails when the J616c is given the
M4 Pro's chip.
allanclempe added a commit to allanclempe/omarchy-m4 that referenced this pull request Nov 18, 2026
aurora 7.3.0.aurora1-40 runs the J616c with the xART guard off
(diag/xart_guard 0), although T6041's profile in aurora-silicon/linux's
current source has guard 1. Touch ID was tried there only with the guard on:
keybag creation, enrolment, matching and matching after a restart, with no
refusal. So the owner's agreement writes xart_guard=1 beside the other two
options on a J616c, in touchid-enable and in enable-fingerprint for a Mac
agreed before the options existed. Other Macs keep the kernel's level: the
guard blocks matching on M1-family Macs. A later kernel with guard 1 makes
it a no-op.

Tests: both writers on a J616c write the guard; the existing assertions keep
other Macs' line without it.

@scottjones scottjones left a comment •

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

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

Thanks for this. The J616c change looks right to me: the board and SoC check matches the kernel's T6041 profile, the chip id is 00006041 in both places that need it, and disable removes the new key. I ran the omarchy-mac suite at 35a4fe7 and merged with current main (124 ok, integration suite passes), and broke each J616c path on purpose: wrong chip, J616c dropped from the board check, the calibration reseal skipped, the key left out of disable, the chip missing from touchid-root-key. Every one fails the new section.

Thanks for eb80585 too. On the 7.3 kernel we're moving edge Macs to (omacom/omarchy-pkgs#3025), T6041 already gets T6040's profile with the guard on, so the explicit xart_guard=1 is a no-op there, as your commit says. Which tree was aurora1-40 from? Worth a line in Merge Danger saying the guard is on either way.

One small test gap: if the T6041 check is dropped so any apple,j616c passes, the tests still pass. A line checking that a J616c on T6040 reads unsupported:board would close it. Not a blocker.

On timing: this waits for #140, which is now ready for review. If #140 lands with a merge commit, this PR shows only your commits. I approved the CI runs and both pass at eb80585. On today's 7.2 kernel a J616c reads none and nothing changes; on 7.3 it needs the v5 libfprint from omacom/omarchy-pkgs#2737, as your notes say.

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