Skip to content

device/bluejay: keep pinned startup outside the passive package - #565

Merged
derekross merged 3 commits into
derekross:mainfrom
latentoperator:revision/bluejay-bringup-20261109
Nov 12, 2026
Merged

derekross merged 3 commits into
derekross:mainfrom
latentoperator:revision/bluejay-bringup-20261109

Conversation

@latentoperator

@latentoperator latentoperator commented Nov 9, 2026 •

Copy link
Copy Markdown
Contributor

Keep the historical Pixel 6a startup helpers, experimental overlays and kernel configuration under device/bluejay/bringup. Nothing is installed automatically or enabled by a package. The historical config fragment has moved out of the normal device build path.

The README pins source c934e316b and release 7.3.0-rc6-omamobile-bluejay-input-gated0, explains historical overlay-label compatibility and /chosen/bootargs replacement, and provides fail-stop install/undo instructions for the root check and ready services. The instructions snapshot prior files and service states before installation and verify the complete snapshot before undo. Run them against the installed development root, not the RAM recovery root. These helpers intentionally refuse the revised kernel stack.

Merge #562 first: its neutral root-ready marker is the prerequisite. This PR does not include the passive package proposed in #346.

Validation on main 6ee77fb: full tests/all, policy and root-ready fixtures pass; independent review exercised install/restore in private roots. No phone operation, new kernel qualification or package installation is claimed. Fresh published-head CI is tracked below.

Published-head validation: both Policy and Tests (tests/all, every package) passed on 5d5bd6bd77c60457d2c3c217eabdee3e0544e12e. The complete combined local integration also passes tests/all and tests/packages with PYTHONUNBUFFERED unset, matching CI.

@derekross

Copy link
Copy Markdown
Owner

Thanks, Chris. Moving the pinned start-up into device/bluejay/bringup/ and dropping the devmem access is exactly what I asked for. Two notes for the README: these helpers belong to kernel c934e316b (they still use the USI probe hold that the revised #229 drops), and input-controller.dtso replaces the whole bootargs line. I'll merge this after #562, since the ready helper clears its marker.

@derekross

Copy link
Copy Markdown
Owner

Thanks, Chris. Both README notes are in as I asked, the helper clears the same marker #562 writes, and I read the whole thing again since the branch was rewritten: no register access from userspace, nothing in the tree enables a unit, and omamobile-bluejay-check holds the phone to the same rules I'd check by hand (slot B, the guard on every partition, AUTO_WRITE=no, boot-success masked).

One file is in the wrong place, and I should have caught it last time. device/bluejay/kernel.config isn't outside the build: bin/omamobile kernel merges device/<codename>/kernel.config into every non-release build for that device. With this file there, any tree built with DEVICE=bluejay would call itself 7.3.0-rc6-omamobile-bluejay-input-gated0, with no commit suffix: the published kernel, your revised stack, and the other Bluejay owner's builds too. That is exactly the release string your start scripts test to make sure they only run on the tested kernel, and the one the notes use to say where a result came from. The module checksums would still stop the helpers, but the name would no longer identify anything. It also changes CONFIG_SPI_S3C64XX from built-in to a module for every Bluejay development build.

Could you move it to device/bluejay/bringup/kernel.config and add a README line saying it is applied by hand when rebuilding c934e316b? Then it really is "outside", like the rest.

Two small things while you're there, neither blocking: a README line that the overlays don't apply to the published kernel (they name labels only the development tree has), and, before these units ever run again, the runbook will need the lines that copy and enable the check and ready services and the lines that undo that.

I'll merge this after #562.

derekross pushed a commit that referenced this pull request Nov 11, 2026
…encies (#561)

Records the Pixel 6a's persistent desktop on its development kernel, what can and can't be reproduced from published sources, the one slot-success write with its before and after bytes, and the hold on further marking. Lists the split pull requests (#562, #563, #565) and the kernel pull requests the port still waits on. Documentation only.

Co-Authored-By: OpenAI Codex <noreply@openai.com>
latentoperator and others added 3 commits November 11, 2026 12:01
Split experimental services, overlays and probe tools from derekross#543. Retain
the exact old development-kernel pins and explicit enable gates; these
helpers refuse the revised kernel topics. Remove userspace display-register
access and use the descriptor/init trigger contract. The proposed SSH-ready
helper clears the neutral root marker only when USB and sshd are ready.
No new phone run or automatic installation is claimed.

Co-Authored-By: OpenAI Codex <noreply@openai.com>
Name the exact historical kernel source and release that the pinned helpers support. Warn that input-controller.dtso replaces the complete chosen bootargs string, so combining overlays requires inspecting both final DT variants and preserving the selected recovery policy.

Co-Authored-By: OpenAI Codex <noreply@openai.com>
Keep the fixed release string and module choices out of ordinary Bluejay builds. Describe manual use with the exact historical kernel, overlay compatibility and installation and restoration of the matching root gates.

Co-Authored-By: Codex <noreply@openai.com>
@derekross

Copy link
Copy Markdown
Owner

Thanks, Chris. The fragment is now device/bluejay/bringup/kernel.config, nothing reads it during a build, and the README says it is applied by hand when rebuilding c934e316b. Both small notes are in too: the line about the overlays and the published kernel, and the install and undo commands for the check and ready pair. I compared this head with the one I read last time: apart from that rename and the README, every file is unchanged.

One thing in the new install section, not blocking: the install block also copies sddm.service.d/bluejay-check.conf, which makes the display manager require omamobile-bluejay-hardware.service and omamobile-bluejay-gpu.service and wait for /run/omamobile-bluejay-hardware-ready. The same section says it doesn't stage the hardware stack, so with only these five files installed sddm can't start, while the Expect says the boot must show the desktop ready. SSH and the ready helper don't depend on sddm, so the marker still clears and nothing is at risk; it's just that the Expect can't pass as written. Could you either leave the drop-in out of the root-gate set, or change the Expect to say the desktop only comes with the hardware and GPU units? That can ride along with the runbook for the next run.

I'll merge this right after #562.

@derekross
derekross merged commit e163141 into derekross:main Nov 12, 2026
2 checks passed
@latentoperator

Copy link
Copy Markdown
Contributor Author

Acknowledging the nonblocking README point for the next runbook: the root-gate-only install must not imply a working desktop. The SDDM drop-in requires the separately staged hardware/GPU units and hardware-ready marker; installing only the check/ready pair cannot meet that desktop expectation.

The next source follow-up will keep that drop-in out of the root-only snapshot/install/undo set and make its Expect cover the root check, SSH and cleared pending marker. Desktop readiness belongs to the separate hardware/GPU integration. No source correction is included in this comment, and this merged historical runbook must not be treated as a revised-kernel recipe. The new headless stage results and restored historical desktop are reported in #348.

robertzas pushed a commit to robertzas/omarchy-mobile that referenced this pull request Nov 14, 2026
…encies (derekross#561)

Records the Pixel 6a's persistent desktop on its development kernel, what can and can't be reproduced from published sources, the one slot-success write with its before and after bytes, and the hold on further marking. Lists the split pull requests (derekross#562, derekross#563, derekross#565) and the kernel pull requests the port still waits on. Documentation only.

Co-Authored-By: OpenAI Codex <noreply@openai.com>
robertzas pushed a commit to robertzas/omarchy-mobile that referenced this pull request Nov 14, 2026
…lled by nothing (derekross#565)

The Pixel 6a's historical start-up helpers, units, overlays, probe tools and kernel config fragment live under `device/bluejay/bringup/` and `overlays/`; no package installs them, no unit is enabled, and the descriptor lists no overlay, so builds and images for bluejay are unchanged. They belong to the tested kernel `c934e316b` (`7.3.0-rc6-omamobile-bluejay-input-gated0`) and refuse any other. The ready helper clears the pending marker derekross#562's `persistent-development` mode writes, once the USB address and sshd are up; the README carries the next-run gate and the install and undo commands for the check and ready pair, none of them executed yet.

Co-Authored-By: OpenAI Codex <noreply@openai.com>
Co-Authored-By: Codex <noreply@openai.com>
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