F43 Kernal 6.19.13 - Nvidia Drivers Not Driving

Ask Fedora
1

Hey All,

I am one of the many who have to battle the Nvidia driver.

Everything was great until I updated, I tried loading the last kernal but it seems borked there as well now. I am trapped at 1024 resoultion.

My ‘unique’ setup is I have a quadro that I am passing to a VM to handle CAD software for work. I have a 5090 as my main gpu. This all worked like a charm.

I removed nvidia using discover. reboot. Reinstall - waited about 15. Then reboot. Nada.

Uninstalled using discover, reboot, dnf install akmod-nividia. Let build, reboot, still nothing.

Ran dmesg(repeats this message a lot):

[  727.082376] NVRM: The NVIDIA GPU 0000:01:00.0 (PCI ID: 10de:2b85)
               NVRM: installed in this system requires use of the NVIDIA open kernel modules.
[  727.082398] NVRM: GPU 0000:01:00.0: RmInitAdapter failed! (0x22:0x56:897)
[  727.083344] NVRM: GPU 0000:01:00.0: rm_init_adapter failed, device minor number 0
[  727.123807] NVRM: The NVIDIA GPU 0000:01:00.0 (PCI ID: 10de:2b85)
               NVRM: installed in this system requires use of the NVIDIA open kernel modules.
[  727.123829] NVRM: GPU 0000:01:00.0: RmInitAdapter failed! (0x22:0x56:897)
[  727.125317] NVRM: GPU 0000:01:00.0: rm_init_adapter failed, device minor number 0
[  727.179796] NVRM: The NVIDIA GPU 0000:01:00.0 (PCI ID: 10de:2b85)
               NVRM: installed in this system requires use of the NVIDIA open kernel modules.
[  727.179818] NVRM: GPU 0000:01:00.0: RmInitAdapter failed! (0x22:0x56:897)
[  727.181284] NVRM: GPU 0000:01:00.0: rm_init_adapter failed, device minor number 0

After reading that, I uninstalled, reboot, then sudo dnf install akmod-nividia-open. build, reboot.

This is where I have landed and am reaching out to smater people. I have jumped between kernal 6.19.13/6.19.12/6.19.11 and its not working in any.

Ouput for “modinfo -F version nvidia”
580.142

I saw this requested in a few other posts so I included it:

System:
  Kernel: 6.19.13-200.fc43.x86_64 arch: x86_64 bits: 64 compiler: gcc
    v: 15.2.1
  Desktop: KDE Plasma v: 6.6.4 tk: Qt v: N/A wm: kwin_wayland dm: SDDM
    Distro: Fedora Linux 43 (KDE Plasma Desktop Edition)
Machine:
  Type: Desktop System: Micro-Star product: MS-7E12 v: 1.0
    serial: <superuser required>
  Mobo: Micro-Star model: MAG X670E TOMAHAWK WIFI (MS-7E12) v: 1.0
    serial: <superuser required> Firmware: UEFI vendor: American Megatrends LLC.
    v: 1.K1 date: 07/16/2025
Battery:
  Device-1: hidpp_battery_0 model: Logitech MX Vertical Advanced Ergonomic
    Mouse serial: <filter> charge: 55% (should be ignored) status: discharging
CPU:
  Info: 8-core model: AMD Ryzen 9 7950X3D bits: 64 type: MCP arch: Zen 4
    rev: 2 cache: L1: 512 KiB L2: 8 MiB L3: 96 MiB
  Speed (MHz): avg: 2986 min/max: 417/5763 boost: enabled cores: 1: 2986
    2: 2986 3: 2986 4: 2986 5: 2986 6: 2986 7: 2986 8: 2986 bogomips: 67199
  Flags-basic: avx avx2 ht lm nx pae sse sse2 sse3 sse4_1 sse4_2 sse4a
    ssse3 svm
Graphics:
  Device-1: NVIDIA GB202 [GeForce RTX 5090] vendor: ASUSTeK driver: nvidia
    v: 580.142 arch: Lovelace pcie: speed: 32 GT/s lanes: 16 bus-ID: 01:00.0
    chip-ID: 10de:2b85
  Device-2: NVIDIA GP106GL [Quadro P2000] driver: vfio-pci v: N/A
    arch: Pascal pcie: speed: 2.5 GT/s lanes: 4 bus-ID: 16:00.0
    chip-ID: 10de:1c30
  Device-3: Logitech HD Pro Webcam C920 driver: snd-usb-audio,uvcvideo
    type: USB rev: 2.0 speed: 480 Mb/s lanes: 1 bus-ID: 3-5:3 chip-ID: 046d:082d
  Display: wayland server: Xwayland v: 24.1.10 compositor: kwin_wayland
    driver: N/A display-ID: 0
  Monitor-1: Unknown-1 res: 1024x768 hz: 60 dpi: 96 diag: 338mm (13.3")
  API: EGL v: 1.5 platforms: device: 0 drv: swrast surfaceless: drv: swrast
    wayland: drv: swrast x11: drv: swrast inactive: gbm
  API: OpenGL v: 4.5 vendor: mesa v: 25.3.6 glx-v: 1.4 direct-render: yes
    renderer: llvmpipe (LLVM 21.1.8 256 bits) device-ID: ffffffff:ffffffff
    display-ID: :0.0
  API: Vulkan v: 1.4.341 surfaces: N/A device: 0 type: cpu
    driver: mesa llvmpipe device-ID: 10005:0000
  Info: Tools: api: clinfo, eglinfo, glxinfo, vulkaninfo
    de: kscreen-console,kscreen-doctor gpu: nvidia-settings,nvidia-smi
    wl: wayland-info x11: xdriinfo, xdpyinfo, xprop, xrandr
Audio:
  Device-1: NVIDIA GB202 High Definition Audio driver: snd_hda_intel v: kernel
    pcie: speed: 32 GT/s lanes: 16 bus-ID: 01:00.1 chip-ID: 10de:22e8
  Device-2: NVIDIA GP106 High Definition Audio driver: vfio-pci pcie:
    speed: 2.5 GT/s lanes: 4 bus-ID: 16:00.1 chip-ID: 10de:10f1
  Device-3: Advanced Micro Devices [AMD] Ryzen HD Audio
    vendor: Micro-Star MSI driver: snd_hda_intel v: kernel pcie: speed: 16 GT/s
    lanes: 16 bus-ID: 17:00.6 chip-ID: 1022:15e3
  Device-4: Generic Realtek USB2.0 Audio
    driver: hid-generic,snd-usb-audio,usbhid type: USB rev: 2.0 speed: 480 Mb/s
    lanes: 1 bus-ID: 3-3:2 chip-ID: 3142:0031
  Device-5: Logitech HD Pro Webcam C920 driver: snd-usb-audio,uvcvideo
    type: USB rev: 2.0 speed: 480 Mb/s lanes: 1 bus-ID: 3-5:3 chip-ID: 046d:082d
  API: ALSA v: k6.19.13-200.fc43.x86_64 status: kernel-api
  Server-1: PipeWire v: 1.4.11 status: active with: 1: pipewire-pulse
    status: active 2: wireplumber status: active 3: pipewire-alsa type: plugin
    4: pw-jack type: plugin
Network:
  Device-1: Realtek RTL8125 2.5GbE vendor: Micro-Star MSI driver: r8169
    v: kernel pcie: speed: 5 GT/s lanes: 1 port: e000 bus-ID: 06:00.0
    chip-ID: 10ec:8125
  IF: enp6s0 state: up speed: 1000 Mbps duplex: full mac: <filter>
  Device-2: Intel Wi-Fi 6 AX200 driver: iwlwifi v: kernel pcie:
    speed: 5 GT/s lanes: 1 bus-ID: 0d:00.0 chip-ID: 8086:2723
  IF: wlp13s0 state: down mac: <filter>
  Device-3: Ralink RT5370 Wireless Adapter driver: rt2800usb type: USB
    rev: 2.0 speed: 480 Mb/s lanes: 1 bus-ID: 1-6:3 chip-ID: 148f:5370
  IF: wlp18s0u6 state: down mac: <filter>
  IF-ID-1: tailscale0 state: unknown speed: -1 duplex: full mac: N/A
Bluetooth:
  Device-1: Intel AX200 Bluetooth driver: btusb v: 0.8 type: USB rev: 2.0
    speed: 12 Mb/s lanes: 1 bus-ID: 3-7:5 chip-ID: 8087:0029
  Report: btmgmt ID: hci0 rfk-id: 2 state: up address: <filter> bt-v: 5.2
    lmp-v: 11
Drives:
  Local Storage: total: 8.88 TiB used: 1.03 TiB (11.6%)
  ID-1: /dev/nvme0n1 vendor: Western Digital model: WDS500G3X0C-00SJG0
    size: 465.76 GiB speed: 31.6 Gb/s lanes: 4 serial: <filter> temp: 38.9 C
  ID-2: /dev/nvme1n1 vendor: Intel model: SSDPEKNW010T9 size: 953.87 GiB
    speed: 31.6 Gb/s lanes: 4 serial: <filter> temp: 29.9 C
  ID-3: /dev/sda vendor: Western Digital model: WD80EDAZ-11TA3A0
    size: 7.28 TiB speed: 6.0 Gb/s serial: <filter>
  ID-4: /dev/sdb vendor: Kingston model: SA400S37240G size: 223.57 GiB
    speed: 6.0 Gb/s serial: <filter> temp: 28 C
Partition:
  ID-1: / size: 951.28 GiB used: 163.12 GiB (17.1%) fs: btrfs
    dev: /dev/nvme1n1p3
  ID-2: /boot size: 1.9 GiB used: 613.9 MiB (31.5%) fs: ext4
    dev: /dev/nvme1n1p2
  ID-3: /boot/efi size: 598.8 MiB used: 19.3 MiB (3.2%) fs: vfat
    dev: /dev/nvme1n1p1
  ID-4: /home size: 951.28 GiB used: 163.12 GiB (17.1%) fs: btrfs
    dev: /dev/nvme1n1p3
Swap:
  ID-1: swap-1 type: zram size: 8 GiB used: 0 KiB (0.0%) priority: 100
    dev: /dev/zram0
Sensors:
  System Temperatures: cpu: 53.1 C mobo: 30.8 C
  Fan Speeds (rpm): N/A
Info:
  Memory: total: 32 GiB note: est. available: 30.94 GiB used: 5.32 GiB (17.2%)
  Processes: 350 Power: uptime: 23m wakeups: 0 Init: systemd v: 258
    default: graphical
  Packages: pm: rpm pkgs: N/A note: see --rpm pm: flatpak pkgs: 21
    Compilers: gcc: 15.2.1 Shell: Bash v: 5.3.0 running-in: konsole inxi: 3.3.40
2

That almost sounds like you ended up with the closed-source version of the kernel module (which doesn’t work with newer cards).

What does this give?

modinfo -l nvidia

and after you tried to install akmod-nvidia-open, did you still get the same messages?

3

Okay, I was also was thinking the same thing. That dmesg is after installing akmods-nivida-open,which is the last thing I did. Been at this all day, so I am probably off on my reboot count.

modinfo -l nvidia
NVIDIA
4

And did akmod-nvidia-open actually install successfully? From what I can see, it’s no longer available in the repos.

Anyway, seeing NVIDIA from the modinfo -l command definitely shows you have the closed-kernel version. For open-kernel you’d see Dual MIT/GPL.

I think the problem you have (at least with your initial install of akmod-nvidia) is that the presence of the Quadro GPU (which isn’t compatible with the open-kernel driver) is causing the package to think that it has to install the closed-kernel version of the driver, which isn’t compatible with the 5090.

If you just want the open-kernel version (because you’re only going to use the Quadro as a passthrough to a VM), there should still be a way to override the auto-detection and force the installation of the open-kernel version.

Can you try this:

echo "%_with_kmod_nvidia_open 1" | sudo tee -a /etc/rpm/macros.nvidia-kmod
sudo dnf reinstall akmod-nvidia
5

And did akmod-nvidia-open actually install successfully? From what I can see, it’s no longer available in the repos.

I believe so, I did not seem to get any errors. I did sudo dnf search akmod-nivida before running the install to make sure I called it correctly and it was listed.

Alright, ran the commands and gave a wait and rebooted.

modinfo -l nvidia

NVIDIA

dmesg

[   96.119588] NVRM: The NVIDIA GPU 0000:01:00.0 (PCI ID: 10de:2b85)
               NVRM: installed in this system requires use of the NVIDIA open kernel modules.
[   96.119610] NVRM: GPU 0000:01:00.0: RmInitAdapter failed! (0x22:0x56:897)
[   96.121173] NVRM: GPU 0000:01:00.0: rm_init_adapter failed, device minor number 0

I ran dmesg again but took a scoll on up and this is the first mention of the GPUs being loaded.

[    7.659443] NET: Registered PF_QIPCRTR protocol family
[    7.703052] nvidia: module license 'NVIDIA' taints kernel.
[    7.703056] Disabling lock debugging due to kernel taint
[    7.703058] nvidia: module license taints kernel.
[    7.936440] ieee80211 phy1: rt2x00_set_rf: Info - RF chipset 5370 detected
[    7.942258] ieee80211 phy1: Selected rate control algorithm 'minstrel_ht'
[    7.943592] usbcore: registered new interface driver rt2800usb
[    7.951823] rt2800usb 1-6:1.0 wlp18s0u6: renamed from wlan0
[    8.053563] nvidia-nvlink: Nvlink Core is being initialized, major device number 509
[    8.053571] NVRM: GPU 0000:16:00.0 is already bound to vfio-pci.
[    8.056787] nvidia 0000:01:00.0: vgaarb: VGA decodes changed: olddecodes=io+mem,decodes=none:owns=none
[    8.059426] RPC: Registered named UNIX socket transport module.
[    8.059428] RPC: Registered udp transport module.
[    8.059429] RPC: Registered tcp transport module.
[    8.059430] RPC: Registered tcp-with-tls transport module.
[    8.059431] RPC: Registered tcp NFSv4.1 backchannel transport module.
[    8.069427] NVRM: The NVIDIA probe routine was not called for 1 device(s).
[    8.069430] NVRM: This can occur when another driver was loaded and 
               NVRM: obtained ownership of the NVIDIA device(s).
[    8.069432] NVRM: Try unloading the conflicting kernel module (and/or
               NVRM: reconfigure your kernel without the conflicting
               NVRM: driver(s)), then try loading the NVIDIA kernel module
               NVRM: again.
[    8.069434] NVRM: loading NVIDIA UNIX x86_64 Kernel Module  580.142  Tue Mar  3 20:04:04 UTC 2026
[    8.146336] nvidia_uvm: module uses symbols nvUvmInterfaceDisableAccessCntr from proprietary module nvidia, inheriting taint.
[    8.321425] nvidia-modeset: Loading NVIDIA Kernel Mode Setting Driver for UNIX platforms  580.142  Tue Mar  3 18:59:57 UTC 2026
[    8.489274] NVRM: The NVIDIA GPU 0000:01:00.0 (PCI ID: 10de:2b85)
               NVRM: installed in this system requires use of the NVIDIA open kernel modules.
[    8.489297] NVRM: GPU 0000:01:00.0: RmInitAdapter failed! (0x22:0x56:897)
[    8.490105] NVRM: GPU 0000:01:00.0: rm_init_adapter failed, device minor number 0
[    8.529687] NVRM: The NVIDIA GPU 0000:01:00.0 (PCI ID: 10de:2b85)
               NVRM: installed in this system requires use of the NVIDIA open kernel modules.
[    8.529705] NVRM: GPU 0000:01:00.0: RmInitAdapter failed! (0x22:0x56:897)
[    8.530619] NVRM: GPU 0000:01:00.0: rm_init_adapter failed, device minor number 0
[    8.530658] [drm] [nvidia-drm] [GPU ID 0x00000100] Loading driver
[    8.567881] NVRM: The NVIDIA GPU 0000:01:00.0 (PCI ID: 10de:2b85)
               NVRM: installed in this system requires use of the NVIDIA open kernel modules.
[    8.567902] NVRM: GPU 0000:01:00.0: RmInitAdapter failed! (0x22:0x56:897)
[    8.568821] NVRM: GPU 0000:01:00.0: rm_init_adapter failed, device minor number 0
[    8.569129] [drm:nv_drm_dev_load [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to allocate NvKmsKapiDevice
6

When I do that, I see only these:

$ dnf search akmod-nvidia
Updating and loading repositories:
Repositories loaded.
Matched fields: name (exact)
 akmod-nvidia.x86_64    Akmod package for nvidia kernel module(s)
Matched fields: name
 akmod-nvidia-390xx.x86_64      Akmod package for nvidia-390xx kernel module(s)
 akmod-nvidia-470xx.x86_64      Akmod package for nvidia-470xx kernel module(s)

Anyway - let’s see if we can see what happened after you tried to set the “_with_kmod_nvidia_open 1” flag and reinstalled:

What timestamp do you see for this file?

ls -al /var/cache/akmods/nvidia/.last.log

Can you see anything in the log about the selection of the open kernel module?

grep "kernel-open" /var/cache/akmods/nvidia/.last.log
7

Here is my search:

Updating and loading repositories:
Repositories loaded.
Matched fields: name (exact)
 akmod-nvidia.x86_64    Akmod package for nvidia kernel module(s)
Matched fields: name
 akmod-nvidia-open.x86_64       Akmod package for nvidia-open kernel module(s)

ls output:

-rw-r--r--. 2 root root 6911491 Apr 26 12:09 /var/cache/akmods/nvidia/.last.log

and grep returned nothing.

8

Sounds like we might have different sets of repos installed. What does dnf repolist show?

(Not that this is necessarily relevant to the issue.)

Also, what about dnf list --installed "*nvidia*" ?

(In case both akmod packages are now installed and this is causing problems.)

Is that the timestamp you expect for your last install attempt (I don’t know what timezone you’re in)?

If so, then it seems like it did rebuild but didn’t get forced to the open version.

Could you also try:

grep "Runtime" /var/cache/akmods/nvidia/.last.log

I think/hope that that will show us if it’s still trying to autodetect based on the installed card.

9

dnf list:

repo id                                                   repo name                                                         
copr:copr.fedorainfracloud.org:codifryed:CoolerControl    Copr repo for CoolerControl owned by codifryed                    
copr:copr.fedorainfracloud.org:phracek:PyCharm            Copr repo for PyCharm owned by phracek                            
fedora                                                    Fedora 43 - x86_64                                                
fedora-cisco-openh264                                     Fedora 43 openh264 (From Cisco) - x86_64                          
google-chrome                                             google-chrome                                                     
mullvad-stable                                            Mullvad VPN                                                       
rpmfusion-nonfree-nvidia-driver                           RPM Fusion for Fedora 43 - Nonfree - NVIDIA Driver                
rpmfusion-nonfree-steam                                   RPM Fusion for Fedora 43 - Nonfree - Steam                        
tailscale-stable                                          Tailscale stable                                                  
updates                                                   Fedora 43 - x86_64 - Updates

Timestamp wise, that would have been about two-ish hours ago. That sounds about when I gave up.

grep:

2026/04/26 12:09:43 akmodsbuild: + echo 'Runtime detection of kmod_nvidia_open'
2026/04/26 12:09:43 akmodsbuild: Runtime detection of kmod_nvidia_open

also thanks for digging into this with me!

10

Ah I see, you have the rpmfusion-nonfree-nvidia-driver repo rather than the full rpmfusion-nonfree. Not necessarily a problem though.

OK, did you try it with these commands? And was that after the timestamp you saw? If so then I guess the dnf reinstall wasn’t enough to force a rebuild.

echo "%_with_kmod_nvidia_open 1" | sudo tee -a /etc/rpm/macros.nvidia-kmod
sudo dnf reinstall akmod-nvidia
11

The timestamp would align with the last time I ran the install pre-posting on the forum.

The timestamp would not align with me running the commands you sent. Which would be closer to 1:00.

12

Ah OK, so I guess that reinstall wasn’t enough to trigger a rebuild with the new parameter set.

I’m inclined to say let’s try uninstalling and then reinstalling? That should also eliminate any weirdness in case you have both akmod-nvidia and akmod-nvidia-open installed.

sudo dnf remove "*nvidia*" -x nvidia-gpu-firmware
sudo dnf install akmod-nvidia xorg-x11-drv-nvidia-cuda

Then after a few minutes check the timestamp on that .last.log file again and grep to see if it has “kernel-open” in it. (Specifically we’re looking for this line below…)

2026/04/24 18:12:41 akmodsbuild: + mv kernel-open kernel
13

okay, ran remove & then install. No reboot yet.

last.log:

-rw-r--r--. 2 root root 2897231 Apr 26 14:04 /var/cache/akmods/nvidia/.last.log

Which aligns with the current time.

grep:

2026/04/26 14:03:55 akmodsbuild: + mv kernel-open kernel

Well, unless there is something else I should do before a reboot. I will go for that next!

14

Looks promising!

15

Sweet beautiful resolution!

I got both of my monitors back up and the GPU passthrough with looking-glass is working as well.

I don’t fully understand, but I assume I had some leftover drivers that was mucking it all up?

When I uninstalled I ran sudo dnf remove nvidia*

I did not have the -x nvidia-gpu-firmware

Regardless! Thank you for the help on holding my hand through this.

Cheers! :clinking_beer_mugs:

16

Fantastic!

If you didn’t specify that then you would have removed nvidia-gpu-firmware, which is needed if you ever need to use nouveau.

Probably doesn’t immediately break things while you are using Nvidia’s own driver, but it might be a good idea to reinstall it now with sudo dnf install nvidia-gpu-firmware.

17

The Quadro p2000 is pasqual architecture and the last driver that will support it is the current 580 version. The RTX 5090 is only supported by the open driver which is built by default when installing akmod-nvidia. Both are supported by the open driver with the 580 driver so for now you are good.

However:
If/when you upgrade to fedora 44 it will be mandatory that you change to the legacy version of the 580 driver (akmod-nvidia-580xx) that will continue support for the Quadro. If you do not do that before you upgrade the system to f44 you will replace the current 580 driver with the newer 595 driver and lose support for the Quadro – at least until you remove the 595 driver and install the 580xx driver in its place.

This is the same situation for users who have the GTX 10xx series and below GPUs

18

I disagree, the older card is used for gpu passthrough so doesn’t use the host system provided nvidia driver, it uses whatever driver is used in the VM.

However he pascal card does break the auto detection in the rpmfusion packaging, using this is the correct fix.

echo "%_with_kmod_nvidia_open 1" | sudo tee -a /etc/rpm/macros.nvidia-kmod

The user should continue to use 595.xx once it’s released.

19

Please remember that you need to use:

akmods –rebuild –force

(and not dnf reinstall akmod-nvidia) in order to take the %_with_kmod_nvidia_open 1 into account. Once the kmod-nvidia generated by the akmod-nvidia, you will have to reboot

20

The open-kernel driver doesn’t support Pascal. So what OP needs (and now has) is:

  • The open-kernel driver on the host, to support the RTX 5090
  • A Pascal-compatible driver within the VM