Skip to content

Route Goodix 53xc fingerprint readers to Dell's proprietary TOD driver - #7993

Open
anupanup2001 wants to merge 7 commits into
omacom:quattrofrom
anupanup2001:goodix-53xc-fingerprint-tod
Open

anupanup2001 wants to merge 7 commits into
omacom:quattrofrom
anupanup2001:goodix-53xc-fingerprint-tod

Conversation

@anupanup2001

Copy link
Copy Markdown

PR: Route Goodix 53xc fingerprint readers to Dell's proprietary TOD driver

Fixes #7991

What

Goodix 27c6:530c / 533c / 538c readers (Dell machines) have no open-source libfprint driver — the goodixmoc id table starts at 5840. On these, the wizard detects the sensor, installs fprintd + stock libfprint, and then fprintd-enroll dies instantly on "No devices available", reported to the user as "Enrollment failed. Please try again." — a failure retrying can never fix (#7991, also the 533c case from #1590).

  • bin/omarchy-hw-fingerprint-tod — hidden hw- helper: true when a Goodix 53xc reader is present, mirroring omarchy-hw-fingerprint's OMARCHY_USB_DEVICES_PATH override for testability. Exact vid:pid match only — the same vendor's 5840+ readers are driven by the open driver and must not be steered toward a closed blob.
  • omarchy-setup-security-fingerprint — after installing packages, ask fprintd-list whether any driver actually claimed the reader before putting a finger on glass:
    • Goodix 53xc present → explain the situation and offer, via gum confirm --default=false, to install Dell's proprietary driver (libfprint-tod + libfprint-goodix-53xc from the AUR, handling the provides+conflicts libfprint dance the script already does for libfprint-git). Closed source installs on a yes, not silently.
    • any other driverless reader → an actionable message pointing at the fprint supported-devices list, instead of a doomed enroll.
    • re-runs after the TOD install skip reinstalling stock libfprint (provider conflict), and the enroll hint switches to press-and-hold since the 53xc is a press sensor, not a swipe sensor — scrubbing a finger it wants held still both fails stages and trips the firmware's overheat lockout (Device disabled to prevent overheating, the same self-protection behind Stop the lock screen overheating the fingerprint reader #7179), which rapid retries extend.
  • omarchy-remove-security-fingerprint — also drops the TOD packages (omarchy-pkg-drop already ignores names that aren't installed).

PAM wiring is untouched: it comes last, only after a print is enrolled and verified, against whatever driver fprintd loaded.

Why this shape

The driver is a closed binary, so the wizard asks instead of installing silently — same line #7667 draws for Lenovo's ELAN blob, applied before enrollment rather than after a bogus verify. Keeping the known-id list to the blob's verified table (530c, 533c, 538c) follows the "no guessing at siblings" approach from #7667; 5840 stays out because the open driver covers it.

Testing

  • test/shell.d/hw-fingerprint-tod-test.sh (new), same fixture pattern as hw-fingerprint-test.sh: flags 530c/533c/538c, rejects 5840/6014/unknown pids, another vendor's 530c, and an empty bus.
  • Full test/shell and test/cli suites pass (the only failures on this machine are three pre-existing files requiring a local omarchy-pkgs checkout; they fail identically on quattro HEAD).
  • Verified end to end on real hardware (Dell Inspiron 5585, Goodix 27c6:530c): the pre-flight fires, the TOD install completes, fprintd claims the device, and enrollment + verify + PAM setup succeed through the unmodified wizard flow.

Copilot AI balanced review requested due to automatic review settings August 24, 2026 04:10

Copilot AI 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.

Pull request overview

Routes supported Goodix 53xc fingerprint readers through Dell’s proprietary TOD driver.

Changes:

  • Adds exact hardware detection and tests.
  • Adds driver preflight, consent, installation, and enrollment guidance.
  • Removes TOD packages during fingerprint teardown.

Tip

If you aren't ready for review, convert to a draft PR.
Click "Convert to draft" or run gh pr ready --undo.
Click "Ready for review" or run gh pr ready to reengage.

Reviewed changes

Copilot reviewed 1 out of 4 changed files in this pull request and generated no comments.

File Description
bin/omarchy-hw-fingerprint-tod Detects supported Goodix 53xc IDs.
bin/omarchy-setup-security-fingerprint Installs and configures the proprietary driver.
bin/omarchy-remove-security-fingerprint Removes TOD packages.
test/shell.d/hw-fingerprint-tod-test.sh Tests exact device matching.

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

@timhanley

Copy link
Copy Markdown

Tested the approach in this PR by hand on a Dell XPS 15 9500 (Goodix 27c6:533c), Omarchy 4.0.4:

  • libfprint-tod 1.95.2+tod1-1 + libfprint-goodix-53xc 0.0.6-2 from the AUR → fprintd-list finds "Goodix Fingerprint Sensor 53xc"; enroll, verify, sudo, polkit and lock screen all work.
  • The current omarchy-setup-security-fingerprint then fails on this machine because it unconditionally installs libfprint-git, which conflicts with libfprint-tod (removing libfprint-tod breaks dependency 'libfprint-tod' required by libfprint-goodix-53xc). So this PR needs Keep a TOD-provided libfprint when enabling fingerprint auth #12833 (or equivalent) alongside it, or re-running setup after the TOD driver is installed will keep failing.
  • Build gotcha: building libfprint-tod failed in g-ir-scanner (No module named 'distutils') when a mise-managed Python 3.12+ was first in PATH. Probably worth a note if the setup builds it via yay.

Written with Claude Opus 5.5 via Claude Code.

omarchybot and others added 4 commits October 13, 2026 10:40
quattro's omacom#7158 replaced setup_lock_fingerprint_pam with a call to omarchy-apply-lock, next to where this branch adds setup_goodix_tod_driver. Keep the new function and drop the removed one.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
The removal script now also drops libfprint-tod and libfprint-goodix-53xc, and this test, added on quattro after the branch was cut, pins the exact package list.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
fprintd-list also fails on a permission or D-Bus error, and the pre-flight read every failure as a missing driver: a reader libfprint drives would be told no driver supports it, or on 53xc hardware offered the proprietary one. Match fprintd's own "No devices available" and leave any other failure to enrollment, as before the pre-flight.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Codex Medium <noreply@openai.com>
The open provider is removed with -Rdd before the AUR build, so a failed build left fprintd installed without libfprint. Reinstall what was removed, with --ask 4 so it also replaces libfprint-tod if that half had installed.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Codex Medium <noreply@openai.com>
@omarchybot

Copy link
Copy Markdown
Collaborator

Reviewed by Claude Opus 5.5 and, as a second opinion, Codex Medium. Codex's two findings were ones I had also reached, so its agreement doesn't count as an independent check. I pushed four commits to this branch; the head is now d4606750d74795decbf3229ae9501e343e914c00.

What was pushed

  • 111751049 merges quattro, which had stopped merging cleanly: Fix fingerprint enrollment and lock-screen recovery #7158 replaced setup_lock_fingerprint_pam with a call to omarchy-apply-lock right next to the new setup_goodix_tod_driver. The merge keeps your function and drops the removed one.
  • 1121a5c5f updates test/shell.d/security-fingerprint-remove-test.sh. That test landed on quattro after this branch was cut and pins the exact package list the removal script drops, so it failed once the TOD packages were added.
  • f73a76a71: the pre-flight read any fprintd-list failure as a missing driver. A permission or D-Bus error would tell a working reader that no driver supports it, or offer the proprietary driver on 53xc hardware. It now acts only on fprintd's own No devices available, and leaves other failures to enrollment as quattro does. fingerprint-package-test.sh covers both cases.
  • d4606750d: the open libfprint provider is removed with pacman -Rdd before the AUR build, so a failed build (like the distutils failure @timhanley hit) left fprintd without libfprint. The setup now reinstalls whatever it removed, with --ask 4 so this also works when libfprint-tod had installed and only the Goodix plugin failed.

What ran

On a disposable Omarchy worker with real pacman, yay and fprintd, and a fake sysfs device 27c6:530c:

  • quattro installs libfprint-git and fails at once with NoSuchDevice and "Enrollment failed. Please try again."
  • This head offers the driver; declining aborts with nothing removed.
  • Accepting removes libfprint-git, builds and installs libfprint-tod and libfprint-goodix-53xc, and keeps fprintd.
  • A forced build failure, and a run where only the plugin failed, both restore libfprint-git with pacman -Dk clean.
  • A re-run does not pull libfprint-git back, and omarchy-remove-security-fingerprint drops all five packages.
  • ./test/cli passes, as do the fingerprint setup, removal, TOD detection, apply-lock and pkg-drop shell tests.

No CI ran on this pull request.

What is not proven

The worker has no Goodix reader, so nothing here shows Dell's driver actually enrolling and verifying a finger. That rests on your Inspiron 5585 and on @timhanley testing the approach by hand on an XPS 15 9500. A confirmation from the author of the fix can't stand as verification, and you also opened #7991. If @m17kea or @timhanley can run Setup > Security > Fingerprint at d4606750d74795decbf3229ae9501e343e914c00 on a 533c (omarchy dev link to a checkout of this branch) and report whether enrollment, verify and sudo work, that is what the maintainer is missing.

What's next

It waits on the maintainer: whether Omarchy should offer a closed AUR driver at all is their call (#7667 draws the same line for Lenovo's ELAN blob), and only they can confirm it on hardware. Two open pull requests overlap in this script without competing for the 53xc fix. #15080 adds a general message for readers fprintd cannot see, after a failed enrollment; #12833 keeps a TOD-provided libfprint for other vendors' blobs. Whichever lands second will need its messages reconciled with this pre-flight.

@greptile-apps

greptile-apps Bot commented Oct 13, 2026 •

Copy link
Copy Markdown
Contributor

RetriggerConfidence Score: 5/5

[Medium impact] The PR appears safe to merge; both previous findings are addressed and no new actionable issue was found.

Summary

Routes Goodix 530c, 533c, and 538c readers to an optional Dell proprietary driver.

  • Fingerprint setup offers Dell’s driver for Goodix 53xc readers.
  • Dell 53xc enrollment tells people to hold a finger still.
  • Fingerprint removal also drops the Dell TOD driver packages.

Diagram

%%{init: {'theme': 'neutral'}}%%
flowchart TD
  A[Detect fingerprint reader] --> B[Ensure required packages]
  B --> C{Driver sees a reader?}
  C -->|Yes| G[Enroll and verify]
  C -->|No| D{Known Goodix 53xc?}
  D -->|No| E[Explain unsupported reader and stop]
  D -->|Yes| F{User accepts Dell driver?}
  F -->|No| H[Stop without swapping packages]
  F -->|Yes| I[Swap to TOD packages]
  I -->|Failure or interruption| J[Restore open driver and stop]
  I -->|Success| K[Reload rules and restart fingerprint daemon]
  K --> L{Driver sees a reader?}
  L -->|No| M[Explain reboot may be needed and stop]
  L -->|Yes| G
  G -->|Success| N[Configure authentication]
Loading

Reviews (2) · Last reviewed commit: "Put the open libfprint back when the Goo..." · Reviewed by Greptile

Comment thread bin/omarchy-setup-security-fingerprint Resolved
The restore only ran when the AUR build returned a failure, so a Ctrl+C or a closed terminal mid-build exited setup with libfprint removed and fprintd unable to start. It now runs from an exit trap held for the whole swap, with HUP, INT and TERM turned into exits so they reach it. Each provider is recorded before it is removed, because bash runs a pending signal trap as soon as the removal returns, before any line after it. The fingerprint package test now walks the driver path itself: declining, accepting, a failed build, an interrupted build, an interrupted removal, and a rerun on the Dell driver.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Co-Authored-By: Codex Medium <noreply@openai.com>
@omarchybot

Copy link
Copy Markdown
Collaborator

Greptile's review of d4606750d raised two points. I checked both against the source, and both were right. I pushed b2a720bc1 to this branch for them. The review was by Claude Opus 5.5, with Codex Medium as a second opinion.

What was pushed

  • Interrupting the driver setup left fprintd without libfprint. The open provider is removed before the AUR build. The restore only ran when that build returned a failure, so a Ctrl+C or a closed terminal mid-build exited with nothing put back. The restore now runs from an exit trap that is held for the whole swap, and HUP, INT and TERM all reach it. Codex found a narrower window in the first version of this fix: a signal that arrives while pacman -Rdd runs is handled before the next line, so each provider is now recorded before it is removed.
  • No test went through setup_goodix_tod_driver. fingerprint-package-test.sh now runs the driver path with stubbed packages. It covers declining, accepting, a failed build, an interrupted build, an interrupted removal, and a rerun with libfprint-tod already installed.

What ran

On a disposable Omarchy worker, the fingerprint setup, removal and TOD detection shell tests and ./test/cli pass at b2a720bc1. Each new interruption test fails when the setup script is reverted to the version before its fix. Codex's last round on the final commit found nothing open. The only check reported on this pull request is Greptile's review.

What it waits on

Everything else in the previous comment still stands. It waits on the maintainer, who has to decide whether Omarchy should offer Dell's closed driver at all. A worker has no Goodix reader, and the author of a fix can't confirm it, so enrollment on a real 53xc has to be confirmed by a maintainer, at b2a720bc1eb656c079b57c8674b0794e8a8a785c. A test of an earlier head no longer counts.

@omarchybot omarchybot removed bug Something isn't working compatibility Make Omarchy work better on everything! labels Oct 13, 2026
@omarchybot

Copy link
Copy Markdown
Collaborator

The bug and compatibility labels are off this pull request because its classification is now unsettled. Claude Opus 5.5 and Codex Medium classified it again, separately, at b2a720bc1eb656c079b57c8674b0794e8a8a785c, and they disagreed. Claude reads it as a hardware compatibility bug: on Goodix 53xc readers the wizard installs a driver that cannot see the sensor, and enrollment always fails. Codex reads it as an enhancement, because the fix adds new behaviour, an opt-in install of Dell's closed driver from the AUR, rather than correcting existing code. Both readings are defensible, so the maintainer decides which it is.

Nothing else has changed since the last comment. No new commits and no review findings have come in, and the code at this head is what that comment describes. It still waits on the maintainer, who has to decide whether Omarchy should offer the proprietary driver, and on a confirmation of enrollment on a real 53xc reader at this head.

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.

omarchy setup security fingerprint fails instantly on Goodix 27c6:530c/533c/538c readers — no open libfprint driver exists

4 participants