Skip to content

J514s: SEP and Touch ID (T6030) - #1691

Open
RyanTheTide wants to merge 5 commits into
integration/aurorafrom
rescue/pr581-j514s-touchid-v2
Open

RyanTheTide wants to merge 5 commits into
integration/aurorafrom
rescue/pr581-j514s-touchid-v2

Conversation

@RyanTheTide

@RyanTheTide RyanTheTide commented Nov 12, 2026 •

Copy link
Copy Markdown
Contributor

Touch ID on the 14" M3 Pro MacBook Pro (J514s, T6030), rebased onto integration/aurora at 78ea426.

This is Adam Harris's series from #581, taken from the latest revision on his branch (adammharris/linux-aurora:feat/t6030-j514s-touchid @ fd840c9). Authorship and his sign-offs are unchanged. The code is his; the rebase applied without conflicts.

Changes

  • t6030-coprocessors.dtsi, t6030-peripheral-buses.dtsi: integration already describes the SEP, its DART and mailbox, and spi2 (all disabled). This adds the SEP DART's DMA window from the J514s ADT (vm-base 0x10000004000, vm-size 0xffff0000, the same BIT(40) aperture as T8122) and the Mesa sensor as spi2's child (CS 0, 8 MHz, CPHA). Everything stays disabled.
  • t6030-j514s.dts: enables the mailbox, spi2 and the sensor. Power is NUB GPIO 33, active high; the data-ready line (AP GPIO 200) is left out and the driver polls, as on J613. Calibration is apple/mesacal-j514s.bin from the Mac's own iBoot system container.
  • soc: apple: sep: a T6030 profile, which is T8122 with J514s's spi2 base (sensor 0x3352, SPI mode 1, variant-5 key store, xART guard level 1, wide DMA mask). Only apple,t6030 + apple,j514s is matched. J516s shares the SoC but its sensor wiring is unchecked, so it stays off.
  • tools/aurora-sep/tests/test_t6030_sep.py: checks the J514s tree against the ADT values, that J516s leaves the SEP and sensor off, and the profile match.

Testing

On this branch (no hardware):

  • test_t6030_sep.py 4/4 and test_t8122_sep.py 4/4 pass with the kernel's dtc.
  • t6030-j514s.dtb and t6030-j516s.dtb build with W=1 and add no warnings over integration (34 before, 34 after, identical).
  • drivers/soc/apple/ builds with W=1 on the Fedora-base + asahi.config release config (CONFIG_APPLE_SEP=m); no warnings in profile.rs or xart_guard.rs.

Hardware results on the same code before the rebase:

  • Adam Harris (@adammharris), J514s, system firmware 27.0, macOS on the same disk: read-only first load, keybag provisioning with the guard protecting all 10 macOS session records and refusing nothing, sensor up 18 ms after power-on, fprintd enrol and match, a match after reboot and after suspend, and macOS Touch ID still working afterwards.
  • @c4software, a second J514s on Omarchy 4.5.0 with the same series applied to linux-aurora 7.2.11.aurora14: read-only first load, provisioning (level 1, 10 records protected, 0 refusals), enrol and match, enrolment surviving a reboot. A suspend retest with no display attached is still open there.

@adammharris, could you boot this branch on your J514s and confirm the SEP attaches and a match works (after a reboot and, if you can, a suspend)? A short result row here is enough. @c4software, a row from your J514s would be welcome too.

Thanks to Adam Harris for the bring-up, to c4software for the second-machine test, and to malik-na for the review on the earlier PR.

The T6030 profile is selected by the SoC compatible; each board device tree decides whether the SEP and the sensor are enabled.

Assisted-by: LLM (Claude:Opus 5.5)

…ouch ID sensor

The SEP, its DART and mailbox, and spi2 are already described (disabled)
in t6030-coprocessors.dtsi and t6030-peripheral-buses.dtsi. Touch ID
needs two things more, as the J514s ADT (system firmware 27.0) records
them:

- The SEP DART only accepts IOVAs in the window iBoot programmed: vm-base
  0x10000004000, vm-size 0xffff0000, the same BIT(40) aperture as on
  T8122. Describe it with apple,dma-range and apple,dma-offset.
- The Mesa sensor sits on spi2, chip select 0, at 8 MHz with CPHA. Add it
  as spi2's child.

Everything stays disabled: apple_sep enables the SEP and its DART when it
loads, and boards enable the mailbox, the bus and the sensor.

Assisted-by: LLM
Signed-off-by: Adam Harris <amh421@icloud.com>
Signed-off-by: Ryan Murray <ryan@aurorasilicon.org>
Enable the SEP mailbox, spi2 and the Mesa sensor (0x3352) on the 14" M3
Pro. Per the J514s ADT, the sensor is powered from NUB GPIO 33, active
high. As on J613, its data-ready line (AP GPIO 200) is left out and the
driver polls. The calibration comes from this Mac's own iBoot system
container as apple/mesacal-j514s.bin.

J516s shares the SoC, but its sensor wiring is unchecked, so it stays
off.

Assisted-by: LLM
Signed-off-by: Adam Harris <amh421@icloud.com>
Signed-off-by: Ryan Murray <ryan@aurorasilicon.org>
The J514s SEP is pre-booted by iBoot and speaks the variant-5 key store,
and its sensor is the same 0x3352 part T8122 carries, in SPI mode 1. So
J514s takes the T8122 profile with its own spi2 base, including the wide
DMA mask and xART guard level 1 from the first boot. Only J514s is
matched; the other T6030 Mac (J516s) has not had its ADT read.

Add a test that checks the J514s device tree against the ADT values and
that J516s leaves the SEP and the sensor off.

Tested on J514s, system firmware 27.0, with macOS on the same disk:
- the SEP attaches warm and advertises 12 endpoints
- the keybag is provisioned; the guard protects all 10 existing macOS
  session records and has refused nothing
- the sensor answers 18 ms after power-on and registers its serial
- enrolment and a match work, and a match still works after a reboot
  and after suspend
- macOS Touch ID still works afterwards

Assisted-by: LLM
Signed-off-by: Adam Harris <amh421@icloud.com>
Signed-off-by: Ryan Murray <ryan@aurorasilicon.org>
@RyanTheTide RyanTheTide added soc:T6030 Hardware discussions for T6030 (M3 Pro) needs hardware test labels Nov 12, 2026
@c4software

Copy link
Copy Markdown

Thanks for the mention. Yes, happy to add a row from my J514s.

My plan is to first finish the 7.2.11 test (the suspend retest with no display attached, on linux-aurora 7.2.11.aurora14 + the series), then build and boot this 7.3 branch and report here.

@RyanTheTide, if you'd rather I skip straight to 7.3, say so and I'll do that instead.

Drafted with an AI assistant (Claude Code); I reviewed it before posting.

@c4software

Copy link
Copy Markdown

The 7.2.11 suspend retest is done: real s2idle suspend (4 min 25 s, nothing attached), and a fingerprint match still works after resume. Details: omacom/linux-aurora#282 (comment)

Moving on to building this 7.3 branch on the J514s now.

Drafted with an AI assistant (Claude Code); I reviewed it before posting.

@c4software

Copy link
Copy Markdown

Row from my J514s (MacBook Pro 14" M3 Pro, T6030) on this branch at 9ef5088, built with Omarchy's linux-aurora PKGBUILD as 7.3.0-0-1.1691-ARCH. Short version: the enclave side carries over cleanly from 7.2.11, but a match does not complete.

Works

  • Boots, warm attach, profile: T6030/J514s, 12 endpoints, touchid: ready.
  • The keybag and enrolment created under 7.2.11.aurora14 + the series are reused as is: no new keybag, Touch ID: restored 1 enrolled identity/identities; enrolment survived reboot, restore: restored, xart_guard: 1, xart_guard_refusals: 0.
  • Wi-Fi and Bluetooth (BCM4388) are fine with the new PCI IDs.

Fails: reading the sensor frame during a match

apple_sep 29e400000.sep: verify: match operation opened (OPERATION_DEVICE_LIST, 1 device(s))
apple_sep 29e400000.sep: verify: match mode 0x88, user 1000, credential length 16, image init length 0xa7, match context generation 2
spi_master spi0: transfer timed out (remaining 0 tx, 1447 rx)
apple-mesa spi0.0: SPI transfer failed: -110
spi_master spi0: failed to transfer one message from queue
spi_master spi0: noqueue transfer failed
apple_sep 29e400000.sep: verify: reading the frame failed (ETIMEDOUT)
apple_sep 29e400000.sep: verify: match operation ended (MATCH_CANCEL)

Same on a second attempt (remaining 0 tx, 1206 rx). fprintd reports verify-unknown-error ("the fingerprint sensor failed (driver status 1)"). The spi2 interrupt does fire (289108000.spi has counts on every CPU), and the spi2/mesa nodes are the same as in the 7.2.11 build, where the same finger matched an hour earlier, including after suspend. The difference I can see is spi-apple.c: this branch has 6c564f2 ("Service the FIFOs from the interrupt handler"), which aurora14 does not. 42ba050 went back to the thread-driven path for transmit-only transfers because of a stall on J613; here it is the receive side that stalls, so my guess is the same M3 issue on a path that commit does not cover. I have not tested a revert.

Things I had to work around on Omarchy (not about this series)

  • Config: CONFIG_EDAC_APPLE_SOC=m in asahi.config needs CONFIG_EDAC, which Omarchy's base config leaves off.
  • GPU: the DRM driver now reports itself as aurora; Omarchy's mesa-aurora (2026-10-31) only knows asahi, so the compositor crashes and there is no display. All of the above was done over SSH.
  • libfprint: /dev/sep-bio is interface version 5, Omarchy's libfprint accepts only 4. I built libfprint 1.94.100 with the patches from aurora-sep-userspace at 2ea454e0 to get as far as the match.
  • With m1n1-aurora 1.6.1.aurora22: 702280000.usb: dwc3-apple: Failed to get reset control, both external DCPs hit the deferred probe timeout ("Type-C display routes unusable: -110"), and the speakers are disabled ("the boot loader passed no amp levels").

Not tested on 7.3: suspend/resume and enrolment. I am going back to 7.2.11 now, but I kept the build tree and can boot this again to try a patch or a revert of the IRQ FIFO engine if that helps.

Drafted with an AI assistant (Claude Code); I reviewed it before posting.

@RyanTheTide

Copy link
Copy Markdown
Contributor Author

Thanks for bringing Touch ID up on the J514s, the enrol/match/suspend results are great.

One change before this goes in. We select behaviour by a hardware property, not by the list of machines it was tested on, and the board .dts is already where we say which machine has which device:

  • drivers/soc/apple/profile.rs:722-726: if has(b"apple,t6030") && has(b"apple,j514s"). The comment says the J516s is left out only because its sensor wiring is unchecked. That is not a hardware reason.
  • drivers/soc/apple/profile.rs:444-450: controller_base: 0x2_8910_8000 is the spi2 address, which the device tree already describes.

Suggested shape: select the T6030 profile on apple,t6030 (or on the SEP/sensor node's compatible), take the sensor controller from the spi node the sensor sits on, and leave the J516s sensor node disabled in its .dts until someone checks it. That way the profile code has no machine list and the DT stays the only per-board switch.

If you want a cross-check on the shared SEP path afterwards, we can run rows on our M1 Pro / M2 / M4 Pro / A18 Pro.

The T6030 profile was picked only when the root compatible also named
the J514s, so the 16-inch M3 Pro (J516s) and any other T6030 Mac had no
profile at all. The SEP, its DART and spi2 are SoC blocks at the same
addresses on every T6030 Mac, so the SoC compatible selects the profile.

Whether a board actually has the sensor wired is described by its device
tree: only the J514s enables sep_mbox, spi2 and the mesa node today, so
nothing binds on the J516s until its board file enables them.

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

Copy link
Copy Markdown
Contributor Author

Pushed 579cfb2 on top: the T6030 profile is now selected by apple,t6030 alone. The SEP, its DART and spi2 sit at the same addresses on every T6030 Mac; whether a board has the sensor wired stays in its device tree (only t6030-j514s.dts enables sep_mbox, spi2 and mesa, so nothing binds on the J516s until its board file does). The kunit picked() cases and test_t6030_sep.py were updated to match; the selection test passes and rustfmt is clean. A J514s boot row on this head is still owed.

integration added a Mac to the Touch ID kunit list since this branch was
based, so the merged list has 17 entries: bump the array length.

Signed-off-by: Ryan <ryan@aurorasilicon.org>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

needs hardware test soc:T6030 Hardware discussions for T6030 (M3 Pro)

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants