old tesla drivers die on 6.10 ( liquorix )
It seems the old NV 470 drivers die on liquorix 6.10 - other then rewriting the NV dkms bindings is there any other way around this or has the end finally come for this trusty K80 workhorse ?
cant insert images with github URLs here, so here is the repo with screens [link] ______________________________________________________________ :: Code :: root@epyc:~# inxi -bGxxz
System: Kernel: 6.9.12-1-liquorix-amd64 arch: x86_64 bits: 64 compiler: gcc v: 13.2.0 Desktop: GNOME v: 46.0 tk: GTK v: 3.24.41 wm: gnome-shell dm: GDM3 Distro: Ubuntu 24.04 LTS (Noble Numbat) Machine: Type: Server System: Supermicro product: Super Server v: 0123456789 serial: <filter> Chassis: type: 17 v: 0123456789 serial: <filter> Mobo: Supermicro model: H11DSi v: 2.00 serial: <filter> UEFI: American Megatrends v: 2.1 date: 02/21/2020 CPU: Info: 2x 24-core AMD EPYC 7401 [MT MCP SMP] arch: Zen speed (MHz): avg: 2000 min/max: 1200/2000 Graphics: Device-1: NVIDIA GK210GL [Tesla K80] driver: nvidia v: 470.256.02 arch: Kepler pcie: speed: 2.5 GT/s lanes: 16 bus-ID: 04:00.0 chip-ID: 10de:102d Device-2: NVIDIA GK210GL [Tesla K80] driver: nvidia v: 470.256.02 arch: Kepler pcie: speed: 2.5 GT/s lanes: 16 bus-ID: 05:00.0 chip-ID: 10de:102d Device-3: ASPEED Graphics Family vendor: Super Micro driver: ast v: kernel ports: active: VGA-1 empty: Virtual-1 bus-ID: 12:00.0 chip-ID: 1a03:2000 Display: server: X.Org v: 21.1.11 with: Xwayland v: 23.2.6 compositor: gnome-shell driver: X: loaded: modesetting,nvidia unloaded: fbdev,nouveau,vesa dri: swrast gpu: ast display-ID: :0 screens: 1 Screen-1: 0 s-res: 1920x1080 s-dpi: 96 Monitor-1: VGA-1 model: HP ZR2240w res: 1920x1080 dpi: 103 diag: 545mm (21.5") API: EGL v: 1.5 platforms: device: 0 drv: nvidia device: 1 drv: nvidia device: 4 drv: swrast gbm: drv: kms_swrast surfaceless: drv: swrast x11: drv: swrast inactive: wayland,device-2,device-3 API: OpenGL v: 4.6.0 compat-v: 4.5 vendor: mesa v: 24.0.9-0ubuntu0.1 glx-v: 1.4 direct-render: yes renderer: llvmpipe (LLVM 17.0.6 256 bits) device-ID: ffffffff:ffffffff Network: Device-1: Intel I350 Gigabit Network vendor: Super Micro driver: igb v: kernel pcie: speed: 5 GT/s lanes: 4 port: 1020 bus-ID: 01:00.0 chip-ID: 8086:1521 temp: 52.0 C Device-2: Intel I350 Gigabit Network vendor: Super Micro driver: igb v: kernel pcie: speed: 5 GT/s lanes: 4 port: 1000 bus-ID: 01:00.1 chip-ID: 8086:1521 Device-3: Intel Ethernet 10-Gigabit X540-AT2 vendor: Inspur Information Industry 10G base-T DP EP102Ti3A driver: ixgbe v: kernel pcie: speed: 5 GT/s lanes: 8 port: N/A bus-ID: 61:00.0 chip-ID: 8086:1528 Device-4: Intel Ethernet 10-Gigabit X540-AT2 vendor: Inspur Information Industry 10G base-T DP EP102Ti3A driver: ixgbe v: kernel pcie: speed: 5 GT/s lanes: 8 port: N/A bus-ID: 61:00.1 chip-ID: 8086:1528 Device-5: Realtek USB 10/100/1G/2.5G LAN driver: r8152 type: USB rev: 3.2 speed: 5 Gb/s lanes: 1 bus-ID: 2-1:2 chip-ID: 0bda:8156 Drives: Local Storage: total: raw: 4.55 TiB usable: 4.55 TiB used: 520.53 GiB (11.2%) Info: Memory: total: 64 GiB available: 62.76 GiB used: 5.25 GiB (8.4%) Processes: 1403 Power: uptime: 29m wakeups: 0 Init: systemd v: 255 target: graphical (5) default: graphical Packages: 1873 pm: dpkg pkgs: 1865 pm: snap pkgs: 8 Compilers: gcc: 13.2.0 Shell: Bash v: 5.2.21 running-in: gnome-terminal inxi: 3.3.34 Back to top |
|||||
Unfortunately, if you want to use the 470 drivers, the best way to get support is through Arch Linux. They have an AUR package with patches to make 6.10 work here: aur.archlinux.org/packages/nvidia-470xx-dkms
Frozen pool distributions like Ubuntu don't backport patches for kernels they don't support, so you're on your own if you need to keep using that distribution. Back to top |
|||||
Just to be clear, a non free proprietary blob like Nvidia tesla no longer supporting current kernels is not a Liquorix issue.
While nvidia does a very good job supporting legacy hardware for a long time, when that support ends, you are much better off either dumping the card and getting an AMD gpu if that's an option, or getting a turing or newer nvidia gpu, which has most of its driver in firmware blobs on the hardware itself. Eventually there will be a consumer level free wrapper layer between the kernel and that blob making driver support less difficult. Given how long this is taking to get working, ti's clear they've hit real obstacles in the process. They referring to nvidia, not the kernel. It's also worth being aware that just because someone hacked out a patch that makes it possible to compile the module for newer kernels does NOT mean that patch is stable, well done, or will work on a newer kernel. What it means is you have a timer started and you should expect support to vanish. Try nouveau and see if it works on that hardware, which is not given, if it does, you should be fine, if not, get ready to replace that card/gpu. Non free unsupporoted EOL drivers are never the problem of a current kernel. It's important to learn what that actually means before buying nvidia gpus. Back to top |
|||||
Thanks all for the replies: replaced the card with a Turing device and 550 drivers to match it
Back to top |
|||||
All times are GMT - 8 Hours
|