Nvidia GPU: Failed to allocate NvKmsKapiDevice

error3 · 2025-10-01 11:47:32

Hi,

today my GTX 1080 stopped working. I plugged the HDMI cable into the motherboard to get onboard graphics. Running wayland.

lspci | grep -i nvidia

01:00.0 VGA compatible controller: NVIDIA Corporation GP104 [GeForce GTX 1080] (rev a1)
01:00.1 Audio device: NVIDIA Corporation GP104 High Definition Audio Controller (rev a1)

uname -a

Linux Odin 6.16.8-arch3-1 #1 SMP PREEMPT_DYNAMIC Mon, 22 Sep 2025 22:08:35 +0000 x86_64 GNU/Linux

dmesg
https://paste.xinu.at/4u9/

After searching the forums I tried downgrading drivers to 570 and 575, both didn't change anything. Now I'm still on 575.
pacman -Q |grep -i nvidia

lib32-nvidia-utils 575.64.05-1
libva-nvidia-driver-git 0.0.13.r7.g9c4a3c7-1
linux-firmware-nvidia 20250917-1
nvidia-dkms 575.64.05-2
nvidia-settings 580.82.09-1
nvidia-utils 575.64.05-2

nvidia-smi

No devices were found

Added nvidia.NVreg_EnableGpuFirmware=0 pcie_aspm=off, no changes observed.

tried downgrading kernel, no changes observed.

nvidia-bug-report.log
https://paste.xinu.at/431v/

Last edited by error3 (2025-10-03 13:59:11)

seth · 2025-10-01 15:01:38

[    2.248352] NVRM: GPU 0000:01:00.0: Failed to copy vbios to system memory.
[    2.248471] NVRM: GPU 0000:01:00.0: RmInitAdapter failed! (0x30:0xffff:1119)
[    2.248520] NVRM: GPU 0000:01:00.0: rm_init_adapter failed, device minor number 0
[    2.248595] [drm:nv_drm_load [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to allocate NvKmsKapiDevice
[    2.248717] [drm:nv_drm_register_drm_device [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to register device

The device doesn't respond at all.

Does the LTS kernel still work?

[    0.000000] DMI: System manufacturer System Product Name/PRIME Z270-A, BIOS 1302 03/15/2018

Is this a notebook? Otherwise take a look at the connections (card seating, dedicated power supply)

There're two nvme's and a sata drive - is there a second OS installed?

Added nvidia.NVreg_EnableGpuFirmware=0 pcie_aspm=off, no changes observed.

Neither is in the dmesg? But ASPM doesn't seem supported anyway.

error3 · 2025-10-01 15:22:18

Hi Seth,

thanks for taking to my issue!

seth wrote:

[    2.248352] NVRM: GPU 0000:01:00.0: Failed to copy vbios to system memory.
[    2.248471] NVRM: GPU 0000:01:00.0: RmInitAdapter failed! (0x30:0xffff:1119)
[    2.248520] NVRM: GPU 0000:01:00.0: rm_init_adapter failed, device minor number 0
[    2.248595] [drm:nv_drm_load [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to allocate NvKmsKapiDevice
[    2.248717] [drm:nv_drm_register_drm_device [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to register device

The device doesn't respond at all.

Does the LTS kernel still work?

Will try LTS kernel next.

seth wrote:

[    0.000000] DMI: System manufacturer System Product Name/PRIME Z270-A, BIOS 1302 03/15/2018
Is this a notebook? Otherwise take a look at the connections (card seating, dedicated power supply)

Desktop PC. All connectors seem firm. Issues happened after manually sending my PC into suspend.
Also found this output from nvidia bug report interesting:

*** /sys/bus/pci/devices/0000:01:00.1/power/runtime_status
*** ls: -r--r--r-- 1 root root 4096 2025-10-01 21:43:52.147903507 +0700 /sys/bus/pci/devices/0000:01:00.1/power/runtime_status
suspended

Maybe it has to do with the supsend?

seth wrote:

There're two nvme's and a sata drive - is there a second OS installed? No just data.
Added nvidia.NVreg_EnableGpuFirmware=0 pcie_aspm=off, no changes observed.
Neither is in the dmesg? But ASPM doesn't seem supported anyway.

That dmesg was from before the edit, here the new one:
https://paste.xinu.at/rKwZvk/

error3 · 2025-10-01 16:21:29

With lts kernel seems to be same problem.

dmesg
https://paste.xinu.at/5OD9zs/

inxi -Gxxxz

Graphics:
  Device-1: Intel HD Graphics 630 vendor: ASUSTeK driver: i915 v: kernel
    arch: Gen-9.5 ports: active: HDMI-A-2 empty: DP-1,HDMI-A-1,HDMI-A-3
    bus-ID: 00:02.0 chip-ID: 8086:5912 class-ID: 0300
  Device-2: NVIDIA GP104 [GeForce GTX 1080] driver: nvidia v: 580.82.09
    arch: Pascal pcie: speed: 8 GT/s lanes: 16 bus-ID: 01:00.0
    chip-ID: 10de:1b80 class-ID: 0300
  Display: wayland server: X.org v: 1.21.1.18 with: Xwayland v: 24.1.8
    compositor: Hyprland v: 0.51.1 driver: X: loaded: modesetting
    alternate: fbdev,intel,vesa dri: iris gpu: i915 display-ID: 1
  Monitor-1: HDMI-A-2 model: Samsung Odyssey G95SC serial: <filter> res:
    mode: 3840x1080 hz: 60 scale: 100% (1) dpi: 82
    size: 1190x340mm (46.97x13.23") diag: 1239mm (48.8") modes: max: 3840x1080
    min: 640x480
  API: EGL v: 1.5 hw: drv: intel iris platforms: device: 0 drv: iris
    device: 1 drv: swrast gbm: drv: iris surfaceless: drv: iris wayland:
    drv: iris x11: drv: iris
  API: OpenGL v: 4.6 compat-v: 4.5 vendor: intel mesa v: 25.2.3-arch1.2
    glx-v: 1.4 direct-render: yes renderer: Mesa Intel HD Graphics 630 (KBL GT2)
    device-ID: 8086:5912
  API: Vulkan Message: No Vulkan data available.
  Info: Tools: api: clinfo, eglinfo, glxinfo, vulkaninfo
    de: kscreen-console,kscreen-doctor gpu: nvidia-settings,nvidia-smi
    wl: wayland-info x11: xdpyinfo, xprop, xrandr

To do further testing I booted live manjaro from USB and even though the

[drm:nv_drm_dev_load [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to allocate NvKmsKapiDevice
[drm:nv_drm_register_drm_device.cold [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to load device

errors don't appear, I still get the other error

NVRM: GPU 0000:01:00.0: Failed to copy vbios to system memory.
NVRM: GPU 0000:01:00.0: RmInitAdapter failed! (0x30:0xffff:1116)
NVRM: GPU 0000:01:00.0: rm_init_adapter failed, device minor number 0

Last edited by error3 (2025-10-01 17:12:13)

error3 · 2025-10-01 18:08:30

Uninstalled nvidia drivers and installed nouveau.

Added nouveau.debug=debug VBIOS=trace.
Corrupt vbios?

dmesg:
https://paste.xinu.at/VFIXpq/

seth · 2025-10-01 18:43:25

[   11.306114] nouveau 0000:01:00.0: enabling device (0000 -> 0003)
[   11.306205] nouveau 0000:01:00.0: NVIDIA GP104 (134000a1)
[   11.306252] nouveau 0000:01:00.0: pci: MSI enabled
[   11.306255] nouveau 0000:01:00.0: bios: trying PRAMIN...
[   11.306272] nouveau 0000:01:00.0: bios: ... not enabled
[   11.306273] nouveau 0000:01:00.0: bios: trying PROM...
[   11.314168] nouveau 0000:01:00.0: bios: 00000000: ROM signature (0000) unknown
[   11.314170] nouveau 0000:01:00.0: bios: image 0 invalid
[   11.314177] nouveau 0000:01:00.0: bios: scored 0
[   11.314178] nouveau 0000:01:00.0: bios: trying ACPI...
[   11.314180] nouveau 0000:01:00.0: bios: trying ACPI...
[   11.314181] nouveau 0000:01:00.0: bios: trying PCIROM...
[   11.314203] nouveau 0000:01:00.0: Invalid PCI ROM header signature: expecting 0xaa55, got 0xac00

today my GTX 1080 stopped working

Anything happened yesterday? Did you enable secure boot or so?
Have you tried to re-seat the GPU (just because)?

https://forums.developer.nvidia.com/t/d … /283329/20

error3 · 2025-10-01 19:47:59

seth wrote:

[   11.306114] nouveau 0000:01:00.0: enabling device (0000 -> 0003)
[   11.306205] nouveau 0000:01:00.0: NVIDIA GP104 (134000a1)
[   11.306252] nouveau 0000:01:00.0: pci: MSI enabled
[   11.306255] nouveau 0000:01:00.0: bios: trying PRAMIN...
[   11.306272] nouveau 0000:01:00.0: bios: ... not enabled
[   11.306273] nouveau 0000:01:00.0: bios: trying PROM...
[   11.314168] nouveau 0000:01:00.0: bios: 00000000: ROM signature (0000) unknown
[   11.314170] nouveau 0000:01:00.0: bios: image 0 invalid
[   11.314177] nouveau 0000:01:00.0: bios: scored 0
[   11.314178] nouveau 0000:01:00.0: bios: trying ACPI...
[   11.314180] nouveau 0000:01:00.0: bios: trying ACPI...
[   11.314181] nouveau 0000:01:00.0: bios: trying PCIROM...
[   11.314203] nouveau 0000:01:00.0: Invalid PCI ROM header signature: expecting 0xaa55, got 0xac00

today my GTX 1080 stopped working

Anything happened yesterday? Did you enable secure boot or so?
Have you tried to re-seat the GPU (just because)?

https://forums.developer.nvidia.com/t/d … /283329/20

Re-seating the GPU did the trick. Weird, PC wasn't touched or moved..
Thanks!

edit: unsure how to mark as solved..

Last edited by error3 (2025-10-01 19:48:44)

seth · 2025-10-01 20:21:12

In doubt it will have achieved a power cycle of the GPUs volatile RAM.

Mark resolved threads by editing your initial posts subject - so others will know that there's no task left, but maybe a solution to find.
Thanks.

error3 · 2025-10-03 13:58:43

seth wrote:

In doubt it will have achieved a power cycle of the GPUs volatile RAM.
Mark resolved threads by editing your initial posts subject - so others will know that there's no task left, but maybe a solution to find.
Thanks.

And problem came back after my PC went into suspend today. Same problem like before. Seems that the GPU cant come back from suspend...
Any ideas?

seth · 2025-10-03 14:11:15

Do you have the journal of that boot?
Does re-seating it revive it again?
Try to disable https://wiki.archlinux.org/title/NVIDIA … P_firmware (that's pretty much the only thing that could survive a state through a reboot)
Edit: you've a 1080, I'm not sure you're even loading a GSP at all - still worth a shot…

Last edited by seth (2025-10-03 14:12:06)

error3 · 2025-10-04 19:44:37

seth wrote:

Do you have the journal of that boot?
Does re-seating it revive it again?
Try to disable https://wiki.archlinux.org/title/NVIDIA … P_firmware (that's pretty much the only thing that could survive a state through a reboot)
Edit: you've a 1080, I'm not sure you're even loading a GSP at all - still worth a shot…

As of now it seems that a simple reboot is enough to reproduce the problem. Even when there was no errors at all in dmesg. I always need to reseat the GPU for it to work again.

Do you still need that specific journal? There have been so many boot cycles it will be very hard for me to find. But with this new information it doesn't seem related to the suspend at all...

I've had GSP disabled throughout this the entire time:

[    0.000000] Linux version 6.15.5-arch1-1 (linux@archlinux) (gcc (GCC) 15.1.1 20250425, GNU ld (GNU Binutils) 2.44.0) #1 SMP PREEMPT_DYNAMIC Sun, 06 Jul 2025 11:14:36 +0000
[    0.000000] Command line: BOOT_IMAGE=/vmlinuz-linux root=UUID=90e582b0-0335-47cd-9c1e-c51e200c27af rw cryptdevice=UUID=83ab44cb-88a3-4f12-8607-7766b9abac61:crypt:allow-discards,no-read-workqueue,no-write-workqueue loglevel=3 quiet nvidia-drm.modeset=1 nvidia_drm.fbdev=1 nvidia.NVreg_EnableGpuFirmware=0 pcie_aspm=off

Last edited by error3 (2025-10-04 19:49:03)

seth · 2025-10-04 20:04:15

You don't even have to suspend the system?
Booting, shutting down and re-booting (also cold?) is enough to lose the GPU until you re-seat it?

Can you disable the IGP?
Does https://wiki.archlinux.org/title/NVIDIA … ersistence help?
Do you configure https://wiki.archlinux.org/title/PRIME# … Management ?

Almost sounds like it's the board, does resetting the CMOS instead of re-seating the GPU "work"?

error3 · 2025-10-05 10:45:49

seth wrote:

You don't even have to suspend the system?
Booting, shutting down and re-booting (also cold?) is enough to lose the GPU until you re-seat it?

Yes, correct.

seth wrote:

Can you disable the IGP?

I'm worried that if I disable IGP, I wont be able to get any video output anymore.

seth wrote:

Does https://wiki.archlinux.org/title/NVIDIA … ersistence help?

So I tried enabling it but it didnt work and I dont see errors in the journal:
sudo nvidia-persistenced --user drailz
ran without returning anything. I checked but the service still disabled:

nvidia-persistenced.service - NVIDIA Persistence Daemon
     Loaded: loaded (/usr/lib/systemd/system/nvidia-persistenced.service; disabled; preset: disabled)
     Active: inactive (dead)

Then tried
sudo nvidia-persistenced

nvidia-persistenced failed to initialize. Check syslog for more details.

dmesg:
https://paste.xinu.at/9aEz/

seth wrote:

Do you configure https://wiki.archlinux.org/title/PRIME# … Management ?

No I have not configured this.

seth wrote:

Almost sounds like it's the board, does resetting the CMOS instead of re-seating the GPU "work"?

So the GPU is seated on the top lanes and covers the battery I would need to remove in order to reset CMOS.
The bottom lanes are free but when I seat the GPU there I always get the errors and cant get it to work at all.

seth · 2025-10-05 13:56:03

sudo systemctl enable --now nvidia-persistenced.service

The BIOS/UEFI doesn't have a reset feature?
You can only use one of the two slots labeled PCIex16

Likewise: does the UEFI even allow you to disable the IGP when there's no GPU slotted?
I'd assume it falls back to that absent a GPU anyway (but of course understand your concerns)

Oh, and sanity check: you don't forget to attach the dedicated 6/8-pin power supply, do you?

error3 · Yesterday 17:33:57

seth wrote:

sudo systemctl enable --now nvidia-persistenced.service
The BIOS/UEFI doesn't have a reset feature?

No, but after finally rtfm I found the CLRTC pin and that is accessible and I can clear CMOS without reseating the GPU that way. But still no picture without reseating.

seth wrote:

You can only use one of the two slots labeled PCIex16

Yea I was referring to the slot located on the bottom of the board its labelled PCIEX16_3. That doesnt give me video output ever, only the one on the top.

seth wrote:

Likewise: does the UEFI even allow you to disable the IGP when there's no GPU slotted?
I'd assume it falls back to that absent a GPU anyway (but of course understand your concerns)
Oh, and sanity check: you don't forget to attach the dedicated 6/8-pin power supply, do you?

I will check tomorrow about the IGP setting. Ive always double checked that hdmi and power cables were attached and firm.

error3 · Today 08:28:53

Hm cant find anything in the manual. I dont think I can disable IGP from BIOS. This Asus z270 is something else.

Arch Linux

#1 2025-10-01 11:47:32

Nvidia GPU: Failed to allocate NvKmsKapiDevice

#2 2025-10-01 15:01:38

Re: Nvidia GPU: Failed to allocate NvKmsKapiDevice

#3 2025-10-01 15:22:18

Re: Nvidia GPU: Failed to allocate NvKmsKapiDevice

#4 2025-10-01 16:21:29

Re: Nvidia GPU: Failed to allocate NvKmsKapiDevice

#5 2025-10-01 18:08:30

Re: Nvidia GPU: Failed to allocate NvKmsKapiDevice

#6 2025-10-01 18:43:25

Re: Nvidia GPU: Failed to allocate NvKmsKapiDevice

#7 2025-10-01 19:47:59

Re: Nvidia GPU: Failed to allocate NvKmsKapiDevice

#8 2025-10-01 20:21:12

Re: Nvidia GPU: Failed to allocate NvKmsKapiDevice

#9 2025-10-03 13:58:43

Re: Nvidia GPU: Failed to allocate NvKmsKapiDevice

#10 2025-10-03 14:11:15

Re: Nvidia GPU: Failed to allocate NvKmsKapiDevice

#11 2025-10-04 19:44:37

Re: Nvidia GPU: Failed to allocate NvKmsKapiDevice

#12 2025-10-04 20:04:15

Re: Nvidia GPU: Failed to allocate NvKmsKapiDevice

#13 2025-10-05 10:45:49

Re: Nvidia GPU: Failed to allocate NvKmsKapiDevice

#14 2025-10-05 13:56:03

Re: Nvidia GPU: Failed to allocate NvKmsKapiDevice

#15 Yesterday 17:33:57

Re: Nvidia GPU: Failed to allocate NvKmsKapiDevice

#16 Today 08:28:53

Re: Nvidia GPU: Failed to allocate NvKmsKapiDevice

Board footer