CubelaPetar/infra-phytron
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Fix Blackwell GPU driver: open kernel modules, bump pin to 595

Browse Source 
The gpu role pinned nvidia-driver-580-server (proprietary kernel
modules). Those install and load cleanly on the RTX PRO 6000, but the
GPU never initialises: NVRM fails at RmInitAdapter (0x22:0x56:897) and
nvidia-smi reports "No devices were found". Blackwell (10de:2bb5)
requires the open kernel modules.

- defaults/main.yml: nvidia-driver-580-server -> 595-server-open.
  The "-open" part is the actual fix and is version independent. 595 is
  the newest -server branch in the 24.04 archive (610 is desktop only);
  projektplan §5.3 asked for the branch to be re-checked and pinned at
  install time, which had not happened yet. Caveat noted in the file:
  595 ships from multiverse, 580/590 from restricted.
- tasks/main.yml: install_recommends: false, keeps nvidia-settings and
  its GTK chain off a headless server (~65 packages).
- manuals/20260714-nvidia-driver-install.md: same command, documents
  the open-module requirement and the failure signature.
- SETUP.md: tick off the completed OS install steps.

Verified on phy-z-srv-gpu01: driver 595.71.05, CUDA 13.2, RTX PRO 6000
Blackwell at 00000001:5C:00.0 with 97887MiB visible, no NVRM errors in
dmesg, single DKMS version registered.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This commit is contained in:
CubelaPetar
2026-08-31 09:30:02 +02:00
co-authored by Claude Opus 5
parent 3160c6ca28
commit e7ef29c5d5
4 changed files with 25 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files
ansible/roles/nvidia_gpu/defaults/main.yml
+7 -3
View File
@@ -1,7 +1,11 @@
---
# Driver pinned explicitly — RTX PRO 6000 (Blackwell) needs >= 580.
# Re-check for a newer branch at install time (projektplan §5).
nvidia_driver_package: nvidia-driver-580-server
# Driver pinned explicitly (projektplan §5: re-check the branch at install time).
# 595 is the newest "-server" branch in the Ubuntu 24.04 archive; 610 is desktop
# only. MUST be the "-open" variant: Blackwell (PCI ID 10de:2bb5) is not
# supported by the proprietary kernel modules — the closed package installs and
# loads fine but fails at RmInitAdapter, so nvidia-smi reports "No devices were
# found". Note 595 lives in multiverse, while 580/590 are in restricted.
nvidia_driver_package: nvidia-driver-595-server-open
# CUDA apt repository keyring (Ubuntu 24.04)
nvidia_cuda_keyring_url: https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2404/x86_64/cuda-keyring_1.1-1_all.deb