[SOLVED] Core Dumping with vulkantools on Hyprland with Nvidia gfx

evakian · 2025-08-30 20:52:29

Hello. I have been trying, and failing, for some time to debug this issue on my own.

I am trying to get Vulkan working correctly on a system with an NVidia graphics card. My window manager is Hyprland, and I am using the linux-hardened kernel. I also have Home Manager installed, and am using the `nixGL` packages to run applications installed through Home Manager correctly. As per the instructions on the wiki, I am trying to verify the successful installation of Vulkan using the `vulkaninfo` command from the `vulkan-tools` package.

However, this command core dumps with the following errors:

Xlib:  extension "GLX" missing on display ":0".
Xlib:  extension "GLX" missing on display ":0".
pci id for fd 9: 10de:2182, driver (null)
kmsro: driver missing
zsh: segmentation fault (core dumped)  vulkaninfo

I will now provide as much information as I can think of:

I have tried disabling GSP firmware by creating `/etc/modprobe.d/nvidia.conf` and adding this line:

options nvidia NVreg_EnableGpuFirmware=0

Output of `lspci -v -nn -d ::03xx`

01:00.0 VGA compatible controller [0300]: NVIDIA Corporation TU116 [GeForce GTX 1660 Ti] [10de:2182] (rev a1) (prog-if 00 [VGA controller])
        Subsystem: Hewlett-Packard Company Device [103c:8556]
        Flags: bus master, fast devsel, latency 0, IRQ 134
        Memory at a3000000 (32-bit, non-prefetchable) [size=16M]
        Memory at 90000000 (64-bit, prefetchable) [size=256M]
        Memory at a0000000 (64-bit, prefetchable) [size=32M]
        I/O ports at 5000 [size=128]
        Expansion ROM at a4000000 [virtual] [disabled] [size=512K]
        Capabilities: <access denied>
        Kernel driver in use: nvidia
        Kernel modules: nouveau, nvidia_drm, nvidia

Output of `pacman -Q | grep nvidia`

lib32-nvidia-utils 580.76.05-1
libva-nvidia-driver 0.0.14-1
linux-firmware-nvidia 20250808-1
nvidia-cg-toolkit 3.1-8
nvidia-dkms 580.76.05-4
nvidia-utils 580.76.05-4

I've tried this with a variety of different packages, such as `nvidia-open-dkms nvidia-utils lib32-nvidia-utils` and `nvidia-open-beta-dkms nvidia-utils-beta lib32-nvidia-utils-beta` to the same results.

Output of `pacman -Q | grep vulkan`

lib32-vulkan-icd-loader 1.4.321.0-1
lib32-vulkan-mesa-layers 1:25.2.1-1
vulkan-headers 1:1.4.321.0-1
vulkan-icd-loader 1.4.321.0-1
vulkan-mesa-layers 1:25.2.1-4
vulkan-tools 1.4.321.0-1
vulkan-validation-layers 1.4.321.0-1

Output of `ls /usr/share/vulkan/icd.d`:

nvidia_icd.json

Output of `VK_LOADER_DEBUG=all vulkaninfo`: https://textbin.net/8n8ysq4ugl

Backtrace of the core dump using GDB:

(gdb) run
Starting program: /usr/bin/vulkaninfo
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/usr/lib/libthread_db.so.1".
Xlib:  extension "GLX" missing on display ":0".
Xlib:  extension "GLX" missing on display ":0".
pci id for fd 9: 10de:2182, driver (null)
kmsro: driver missing

Program received signal SIGSEGV, Segmentation fault.
0x0000000000000000 in ?? ()
(gdb) bt
#0  0x0000000000000000 in ?? ()
#1  0x00007fffef44f116 in ?? () from /usr/lib/libnvidia-eglcore.so.580.76.05
#2  0x00007fffef53bbe2 in ?? () from /usr/lib/libnvidia-eglcore.so.580.76.05
#3  0x00007fffef43be60 in ?? () from /usr/lib/libnvidia-eglcore.so.580.76.05
#4  0x00000001000332d7 in AppQueueFamilyProperties::AppQueueFamilyProperties (this=0x7fffffffca70, inst=..., physical_device=<optimized out>, family_properties=...,
    queue_index=<optimized out>, pNext=<optimized out>) at /usr/include/c++/15.1.1/bits/stl_iterator.h:1058
#5  AppGpu::AppGpu (this=<optimized out>, inst=..., id=<optimized out>, phys_device=<optimized out>, show_promoted_structs=<optimized out>, this=<optimized out>, inst=...,
    id=<optimized out>, phys_device=<optimized out>, show_promoted_structs=<optimized out>) at /usr/src/debug/vulkan-tools/Vulkan-Tools/vulkaninfo/./vulkaninfo.h:1690
#6  0x000000010002bf37 in main (argc=<optimized out>, argv=<optimized out>) at /usr/src/debug/vulkan-tools/Vulkan-Tools/vulkaninfo/vulkaninfo.cpp:1308

Any assistance would be apprciated, thanks

Last edited by evakian (2025-09-02 18:56:40)

seth · 2025-08-31 07:58:01

You're somehow crashing in xwayland

DISPLAY= vulkaninfo --summary

also please post your Xorg log, see https://wiki.archlinux.org/title/Xorg#General fro where to find it.

StarWolf3000 · 2025-08-31 08:15:54

Post the contents of

/usr/share/vulkan/icd.d/nvidia_icd.json

(I have reasons to ask for this)

Last edited by StarWolf3000 (2025-08-31 08:17:24)

tekstryder · 2025-08-31 14:01:01

StarWolf3000 wrote:

I have reasons

Are your "reasons" this unsubstantiated claim later debunked by an nVidia developer?

Looks more like a misconfigured setup.

evakian wrote:

using the `nixGL` packages

If you're committed to whatever this is, did you try:

nixVulkanNvidia vulkaninfo --summary

• https://github.com/nix-community/nixGL

evakian · 2025-09-01 14:01:15

seth wrote:

You're somehow crashing in xwayland
DISPLAY= vulkaninfo --summary
also please post your Xorg log, see https://wiki.archlinux.org/title/Xorg#General fro where to find it.

Ah, this was a giveaway.

It appears in the my Home-Manager configuration I had the `xwayland.enable = true;` option turned on, meaning that `Xwayland` was being installed and that the first `Xwayland` in my PATH was from `~/.nix-profile/bin/Xwayland` rather than `/usr/binXwayland`.

I believe that this was causing Hyprland (which is also the system version, from `/usr/bin/Hyprland`) to use the `~/.nix-profile/bin/Xwayland` when launching Xwayland, which I don't think is configured correctly.

Once I changed this to `xwayland.enable = false;`, and rebooted, I'm now able to see `vulkaninfo` run cleanly.

tekstryder wrote:

StarWolf3000 wrote:
I have reasons
Are your "reasons" this unsubstantiated claim later debunked by an nVidia developer?
Looks more like a misconfigured setup.
evakian wrote:
using the `nixGL` packages
If you're committed to whatever this is, did you try:
nixVulkanNvidia vulkaninfo --summary
• https://github.com/nix-community/nixGL

No, I'm trying to focus on confirming that the base operating system had everything set up correctly before I try verifying all the functionality of the Home Manager stuff. I wouldn't expect the Archlinux forums to support Home Manager problems anyways, I only mentioned that to provide more context for my system. I was assuming (I guess incorrectly) that none of the stuff installed through Home Manager would conflict with the base operating stuff.

Thanks for the help, all.

seth · 2025-09-01 14:35:58

\o/
Please always remember to 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.

Arch Linux

#1 2025-08-30 20:52:29

[SOLVED] Core Dumping with vulkantools on Hyprland with Nvidia gfx

#2 2025-08-31 07:58:01

Re: [SOLVED] Core Dumping with vulkantools on Hyprland with Nvidia gfx

#3 2025-08-31 08:15:54

Re: [SOLVED] Core Dumping with vulkantools on Hyprland with Nvidia gfx

#4 2025-08-31 14:01:01

Re: [SOLVED] Core Dumping with vulkantools on Hyprland with Nvidia gfx

#5 2025-09-01 14:01:15

Re: [SOLVED] Core Dumping with vulkantools on Hyprland with Nvidia gfx

#6 2025-09-01 14:35:58

Re: [SOLVED] Core Dumping with vulkantools on Hyprland with Nvidia gfx

Board footer