Repository navigation
Conversation
|
hive/awaiting-ci: CI pending — sweep will merge when green. 🐝 Hive Agent: |
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: castrojo, danathar The full list of commands accepted by this bot can be found here. Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
|
Stand-in review at 65ce85c: correct, and good to merge. Checked against systemd 259's shipped rules:
Worth knowing, not blockers:
Not an approval: this is Doug's PR, it's printer work, and common needs two approvals. |
|
@Danathar this PR is back in your court ( Failing checks:
Reply here or push a new head to this branch (same sign-off); CI and the review swarm re-run automatically and the PR returns to the normal path. Queued in your relay fix queue (1 waiting): fixes are handed out before new work as soon as you have capacity. |
|
Status: the red checks on
I'll re-run once #1668 lands; nothing else in this PR needs to change. 🐝 Hive Agent: |
65ce85c to
aefee45
Compare
|
Follow-up: #1668 has landed, and 🐝 Hive Agent: |
|
Re-review after the rebase (aefee45): the rule and |
There was a problem hiding this comment.
This looks correct to me.
What I checked against the tree at the merge base:
- Match.
ENV{ID_USB_INTERFACES}=="*:0701??:*"uses the interface-class glob systemd's50-udev-default.rulesalready uses to setGROUP="lp"on USB printers. The rule fires only onDEVTYPE=="usb_device", which is the/dev/bus/usb/BBB/DDDnode that libusb-based printer apps open. - Ordering. The rule is
70-…, so it runs afterusb_idfills inID_USB_INTERFACESand before73-seat-late.rulesturns theuaccesstag into an ACL. The existing guard testuaccess rules sort before systemd's 73-seat-late.rulesintests/test_udev_rules.batscovers that ordering for the new file automatically. - Scope. The rule has no
MODEorGROUP, solpkeeps its access and only the active seat user gains an ACL. The new negative test pins this. - Wiring.
tests/test_udev_rules.batsis already run fromJustfile:55, so the three new tests execute injust test.
For the human reviewer: the rule is in system_files/shared/, so it also ships to Aurora. It grants the local active-session user raw write access to any USB printer-class device. That is the intent of #1681, and it follows the same seat-ACL pattern as the other uaccess rules in this directory.
Confidence: 5/5 (safe)
— hive: agent=reviewer backend=claude model=claude-opus-5-5 effort=medium claude=2.1.325
|
Hive status — updated in place; edits do not notify. Full history is in the Activity Monitor. Review findings from This looks correct to me. What I checked against the tree at the merge base:
For the human reviewer: the rule is in |
|
The red This PR changes only a udev rule and its bats tests, and its required checks ( 🐝 Hive Agent: |
Summary: Printer apps that run as the logged-in user can see a USB printer but can't send it anything to print, because only the system's printing group may write to it and nobody is in that group by default. This change tells the system to give whoever is logged in at the computer permission to use USB printers too, the same way it already handles security keys and similar devices. It is the system half of letting the setup app (ChairLift) hand a USB printer to a printer app.
bluefin-common PR
What does this change?
Adds
70-usb-printer-uaccess.rules, which tags USB devices with a printer-class (0701) interface withTAG+="uaccess", plus bats tests for it.Why?
Fixes #1681
systemd's
50-udev-default.rulesleaves USB printer nodesroot:lp0664, hplip's56-hpmud.rulesdoes the same for HP printers, andlphas no members on a default install. A rootless printer application (ChairLift passes the printer in, projectbluefin/chairlift#763) can open the node read-only but cannot write a job to it.uaccessmakes systemd-logind add a seat ACL for the active user, so the rule sets noMODEand noGROUP:lpkeeps its access and nothing becomes world-writable. The file is numbered70-so it sorts before73-seat-late.rules, where the tag is applied (the existing ordering test covers this).Tests added in
tests/test_udev_rules.bats: the file exists; it matchesSUBSYSTEM=="usb",ENV{DEVTYPE}=="usb_device",ENV{ID_USB_INTERFACES}=="*:0701??:*"and carriesTAG+="uaccess"; and it sets noMODEorGROUP. The existing ordering and world-accessible-mode checks run over the new file too.Not tested: I did not run bats locally (CI runs it), and I had no USB printer to check
getfacl /dev/bus/usb/<bus>/<device>showsuser:<you>:rw-after login. A reviewer with a USB printer can confirm that on a built image.How this PR merges
Prow requests two reviews from
OWNERS. A reviewer who is not the authorcomments
/lgtm(removed on every push), anOWNERSapprover comments/approveor approves in GitHub, and Common also needs 2 approving GitHubreviews. With green checks, Prow puts the PR in the merge queue.
/holdpauses it. See how issues and PRs work here.
Checklist
fix:,feat:,docs:,ci:,refactor:, etc.)just checkpasses — Justfile unchanged; left to CIpre-commit run --all-filespasses — left to CIAGENTS.md/docs/SKILL.md/docs/skills/links remain validgh run list --repo projectbluefin/common --limit 5AI attribution
The commit carries
Assisted-by: Claude Opus 5.5 via Claude Codeand aCo-authored-bytrailer.🤖 Generated with Claude Code
— hive: agent=scanner backend=claude model=claude-opus-5-5
🐝 Hive Agent:
contributor| SHA:65ce85c