[satchlj]

Aurora already has frame gain/exposure reporting from Ryan’s original #689. Port the remaining section-table entry bound and short-header/control-gate regression tests. Make the helper enforce the same hardware-profile gate as Aurora’s existing caller, and retain newer T6040/P010 support. T8122 metadata units/layout remain unqualified and gain reporting stays gated off.

Port of omacom/linux-aurora#349. The original PR remains untouched.

Carried dependency: ef2275f384be (carried commit ba71c0044845), media: apple: isp: report the frame gain and exposure, original author Ryan Murray: carried unaltered from omacom/linux-aurora. Its complete patch, message, authorship and dates are preserved; the separate 7.3 adaptation retains Aurora’s existing capture support.

@RyanTheTide

This carries Ryan Murray's additional metadata bounds and regression coverage from omacom/linux-aurora unaltered, because Aurora doesn't have it yet and this change builds on it. If you'd rather land it separately or differently, say so and I'll rework this.

Validation: arm64 defconfig + asahi.config, Image/modules/DTBs with BTF and KUnit enabled, LLVM 18/Rust 1.85.0, passed at stack top #349 (76db72372939), which includes this branch. The validated integration base is b500668. QEMU loaded the test-bearing modules successfully with no additional failures. The combined run retains four pre-existing SEP store failures (plus their suite aggregate), reproduced on the unmodified Aurora base. The unmodified pinned base also crashes in apple-av-audio-lifetime/av_lifetime_devres_order; that unrelated audio suite is explicitly excluded. All tests added by these ports run through KUnit debugfs on Cortex-A76. This is not a fully green or complete-tree KUnit run. The additional short-header/control-gate KUnit coverage ran in QEMU. The latter first caught the missing helper-side check; the production helper was corrected while retaining the caller gate and unchanged assertions. No 7.3 hardware boot or deployment has been performed.

[satchlj]

@DjDeveloperr @RyanTheTide ready for your regression testing. It hasn't been boot-tested on J613 with 7.3 yet; I'll post the result here when I have it. Still a draft until the camera gain is checked on J613 hardware.

[satchlj]

J613 boot result on 7.3 (rubra, M3 MacBook Air 13", macOS 26.6.2 / 25G83 firmware), 2026-11-05.

• Kernel: integration/aurora 11eccbed with all 17 J613 PRs merged (soc: apple: sep: preserve unreadable and partial identity state #962–media: apple: isp: bound exposure metadata and keep controls gated #978), plus HID: dockchannel-hid: write each frame to the FIFO under one lock #1010 (the MTP frame-lock fix). Boot chain: stage2 from the m1n1 Thunderbolt display on M1 Pro/Max (t600x), stacked on #8 #64 content, U-Boot with arm64: dts: apple: t8140/j700: keyboard, trackpad and USB-C topology (reworked, 5 patches) #30/soc: apple: sep: T6000 (M1 Pro) support, plus SPI CS timing and attach fixes #26, then Limine.
• Mesa: aurora-wip 4371cf86 + mesa cpufreq: apple: add bounded M6 T8152 CPU scaling #161–#163.
• Automatic checks passed:
  • boots to the desktop;
  • GPU: DRM driver aurora, Apple M3 (G15G) for GL and Vulkan, glmark2 1113 (1011 on 7.2.11), no GPU errors;
  • internal panel up;
  • keyboard and trackpad present with no MTP crash;
  • camera node, Wi-Fi, Bluetooth, audio cards and headphone jack;
  • Touch ID policy readback (identity_source=dt);
  • speakers stay off as intended (our stage2 passes no apple,speaker-config).
• Not checked yet, since they need me at the machine: the fingerprint match, an external display, and suspend.

For this PR: the camera node comes up. Frame-gain image quality hasn't been judged yet.

[RyanTheTide]

The net change against integration/aurora is small: the section-table bound in isp_meta_read_exposure(), the meta_exposure gate in isp_init_exposure_ctrls(), and the two KUnit additions (37 lines). The history doesn't match that:

• 16e17fc ("Prepare the original context ...") removes about 800 lines of integration's ISP code, including the T6040/P010 support and the existing frame gain/exposure reporting.
• ba71c00 then re-adds "media: apple: isp: report the frame gain and exposure", which integration already has.
• d83a42a composes the two back again.

The intermediate commits don't build to a working driver for T6040/T8140, so bisecting breaks. Also, 16e17fc's subject has no subsystem prefix. Please drop those three commits and carry the net change as a single commit on integration/aurora (for example media: apple: isp: bound the exposure section entry and gate the controls in the helper), keeping the original author credit in the message if you want it.

Testing: T8122 keeps meta_exposure off, so this change doesn't need a J613 gain check. The new size check does run on T8140 and T6040, though. Please run isp_test_meta_exposure and isp_test_exposure_control_gate under KUnit. Those two boards also need a capture showing that V4L2_CID_GAIN and V4L2_CID_EXPOSURE_ABSOLUTE still update.

[satchlj]

Replaced the remove/re-add history with one media: apple: isp: commit carrying only the net bounds check, helper gate and two KUnit additions on integration/aurora. It is independent of the ISP preview PR. Both isp_test_meta_exposure and isp_test_exposure_control_gate passed under arm64 QEMU KUnit (2/2). T8122 keeps metadata exposure disabled, so no J613 gain check is claimed or needed. T8140 and T6040 captures showing live gain and absolute-exposure updates remain owed; I don't have either board.

[DjDeveloperr]

Codex (GPT-6.1 Sol), reviewing on @DjDeveloperr’s behalf.

The current net bounds/helper-gate change and history cleanup look good; the posted two targeted KUnit passes cover the requested source regressions. T8122 has metadata exposure disabled, so a J613 camera test cannot close the live-control gate. Keep draft pending capture on T8140/Neo and T6040 showing V4L2_CID_GAIN and V4L2_CID_EXPOSURE_ABSOLUTE still update with no new ISP/DART errors. Hardware coordination request sent to ryan-orch-regress; frames must remain local.

[RyanTheTide]

Hardware results (unattended run).

Boards: J700 (T8140) and J616s (T6040).
Baseline: integration/aurora 84d091f.
Test kernel: 84d091f + this PR at 8f9ffa3, same config as the baseline.

Both boards pass, with numbers identical to the baseline:

• 8 streams (30/15 fps × exposure_dynamic_framerate 0/1 × power_line_frequency 50/60 Hz), 150 frames each, none lost. 30.06–30.16 fps at 30, 15.01–15.04 fps at 15.
• gain and exposure_time_absolute are read-only and volatile; writes return EACCES. Read from a second fd during streaming, both follow the frames and stay inside the control range: gain 260–4168 (J700) and 260–3123 (J616s), exposure 83–665 (100 µs units). With dynamic framerate off, exposure never went past the frame interval.
• New ISP errors 0, DART faults 0 (baseline 0 and 0).
• KUnit on arm64 QEMU: ISP 39/39, apple-isp-v4l2 11/11 including isp_test_exposure_control_gate.

Not covered: image quality, T8122.

[DjDeveloperr]

Codex (GPT-6.1 Sol), reviewing on @DjDeveloperr’s behalf.

The remaining gate is now satisfied by Ryan's posted hardware result #978 (comment): unchanged head8f9ffa30 merged into fleet63/84d091f7 on J700/T8140 and J616s/T6040, same baseline config, one-shot/default untouched. On each board eight150-frame streams lose no frames; read-only/volatile gain and absolute exposure update within range, writes return EACCES, and no new ISP/DART faults. Reported arm64 QEMU suites pass39/39 and11/11, including the targeted gate/bounds cases. Existing exact-head host CI is green.

Re-reviewed the current net2-file change: the section-table entry is bounded before its read, the helper preserves the existing hardware-profile gate, and the single-commit history retains newer T6040/P010 support. No blocking source/review findings remain. Ready to leave draft and merge this bounded metadata fix. These are attributed unattended control/metadata results, not visual image quality, T8122 metadata support, or physical input testing.