feat: unify GPU permissions across deployments - #183
Merged
Conversation
MioYuuIH
marked this pull request as draft
July 28, 2026 02:31
Collaborator
There was a problem hiding this comment.
@MioYuuIH We provide deb package for setup kdf and dri permission, do we have a specific reason why we don't use this ?
sudo apt update
wget https://repo.radeon.com/amdgpu/30.30.4/ubuntu/pool/main/a/amdgpu-insecure-instinct-udev-rules/amdgpu-insecure-instinct-udev-rules_30.30.4.0-2341068.24.04_all.deb
sudo apt install ./amdgpu-insecure-instinct-udev-rules_30.30.4.0-2341068.24.04_all.deb
Contributor
Author
There was a problem hiding this comment.
Done. Replaced custom udev rules with AMD's official package and verified it on both GPU nodes.
kerwtsai-amd
requested changes
Jul 28, 2026
kerwtsai-amd
marked this pull request as ready for review
July 29, 2026 03:27
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
renderGidand GPU-specificsupplementalGroupsfrom Helm, Hub, Spawner, installer overlays, and generated deployment artifactsamdgpu-insecure-instinct-udev-rulespackage30.30.4.0-2341068.24.04/dev/kfdand DRMrenderD*receive mode0666;card*remains under the normal system policysingleuser.fsGid: 100strictly for notebook storage ownershipPackage deployment
Infrastructure contract
The AMD device plugin and labeller remain separate cluster prerequisites. Device-plugin allocation controls which containers receive GPU devices; AUPLC only requests
amd.com/gpuand does not inject Linux groups.runtime verification
Tested directly on both Ready GPU nodes with a non-root UID
12345process carrying only GID100:/dev/kfdandrenderD*at mode0666were accessiblecard*remainedroot:video 0660and was intentionally inaccessiblerocminfosucceeded onaup-395-23(gfx1151) andaup-sha-nv20(gfx1200)amd.com/gpureceived no GPU device nodesVerification
Known baseline
The complete Hub test sweep still has three pre-existing collection failures when local test stubs do not provide
jupyterhub.scopes; the changed Hub configuration and GPU Spawner tests pass.