Hello everyone! I’ve spent a really long time trying to figure out this issue and I figured it would do me good to actually try and talk to someone about it.
I did a system upgrade not too long ago and have recently noticed that my GPU hardware acceleration is completely turned off. I have an NVIDIA GeForce GTX 1050 Ti GPU and I use the nvidia-580xx-dkms drivers-- as far as I know, this is standard since the Pascal family is no longer supported by the nvidia-open package. vulkaninfo returns this:
'ERROR: [Loader Message] Code 0 : setup_loader_term_phys_devs: Failed to detect any valid GPUs in the current config
ERROR at /usr/src/debug/vulkan-tools/Vulkan-Tools/vulkaninfo/./vulkaninfo.h:249:vkEnumeratePhysicalDevices failed with ERROR_INITIALIZATION_FAILED
and checking my KDE system settings does indeed confirm that my graphics processor is llvmpipe and not my GPU. Anecdotally, I can verify that most games run very slowly compared to before the upgrade. hyfetch correctly gives me my GPU information but OpenGL & Vulkan don’t seem to register it.
The last time I did an upgrade was admittedly quite a while ago, (probably last year) but even back then I was still using the nvidia-580xx-dkms package. I do not have the nomachine package or use the nomodeset kernel parameter. I’ve tried a whole host of things, including switching to nvidia-open or even using nouveau but neither of those change anything. Basically, I have been banging my head against the wall. I’m not a super expert on Linux so I don’t doubt the possibility that I have just misread some trivial step. Any help on this would be greatly appreciated
tl;dr My Pascal-family Nvidia GPU is not detected after an upgrade even though I have the 580xx drivers installed.
journalctl -b -p err doesn’t have much interesting. journalctl -b -gnvidia gives this:
Jun 01 19:32:43 theep dracut-cmdline[221]: Using kernel command line parameters: nvidia_drm.modeset=1 BOOT_IMAGE=/boot/vmlinuz-linux root=UUID=274b571f-99e0-46a8-a60f-ae1610f106a5 rw loglevel=3 quiet
Jun 01 19:32:43 theep kernel: nvidia-nvlink: Nvlink Core is being initialized, major device number 237
Jun 01 19:32:43 theep kernel: nvidia 0000:01:00.0: vgaarb: VGA decodes changed: olddecodes=io+mem,decodes=none:owns=io+mem
Jun 01 19:32:43 theep kernel: NVRM: loading NVIDIA UNIX x86_64 Kernel Module 580.159.03 Fri Apr 24 06:16:47 UTC 2026
Jun 01 19:32:44 theep kernel: nvidia-modeset: Loading NVIDIA Kernel Mode Setting Driver for UNIX platforms 580.159.03 Fri Apr 24 05:49:32 UTC 2026
Jun 01 19:32:44 theep kernel: nvidia_uvm: module uses symbols nvUvmInterfaceDisableAccessCntr from proprietary module nvidia, inheriting taint.
Jun 01 19:32:44 theep kernel: [drm] [nvidia-drm] [GPU ID 0x00000100] Loading driver
Jun 01 19:32:44 theep kernel: caller _nv000799rm+0x11e/0x255 [nvidia] mapping multiple BARs
Jun 01 19:32:44 theep systemd-modules-load[199]: Inserted module 'nvidia_uvm'
Jun 01 19:32:44 theep kernel: [drm] Initialized nvidia-drm 0.0.0 for 0000:01:00.0 on minor 1
Jun 01 19:32:44 theep kernel: nvidia 0000:01:00.0: vgaarb: deactivate vga console
Jun 01 19:32:44 theep kernel: fbcon: nvidia-drmdrmfb (fb0) is primary device
Jun 01 19:32:44 theep kernel: nvidia 0000:01:00.0: [drm] fb0: nvidia-drmdrmfb frame buffer device
Jun 01 19:32:46 theep kernel: input: HDA NVidia HDMI/DP,pcm=3 as /devices/pci0000:00/0000:00:01.0/0000:01:00.1/sound/card1/input18
Jun 01 19:32:46 theep kernel: input: HDA NVidia HDMI/DP,pcm=7 as /devices/pci0000:00/0000:00:01.0/0000:01:00.1/sound/card1/input19
Jun 01 19:32:46 theep kernel: input: HDA NVidia HDMI/DP,pcm=8 as /devices/pci0000:00/0000:00:01.0/0000:01:00.1/sound/card1/input20
Jun 01 19:32:46 theep kernel: input: HDA NVidia HDMI/DP,pcm=9 as /devices/pci0000:00/0000:00:01.0/0000:01:00.1/sound/card1/input21
Jun 01 19:32:48 theep kernel: NVRM: API mismatch: the client 'Xorg' (pid 726)
NVRM: has the version 580.159.04, but this kernel module has
NVRM: the version 580.159.03. Please make sure that this
NVRM: kernel module and all NVIDIA driver components
NVRM: have the same version.
Jun 01 19:32:48 theep kernel: NVRM: API mismatch: the client 'Xorg' (pid 726)
NVRM: has the version 580.159.04, but this kernel module has
NVRM: the version 580.159.03. Please make sure that this
NVRM: kernel module and all NVIDIA driver components
NVRM: have the same version.
Jun 01 19:32:48 theep kernel: NVRM: API mismatch: the client 'Xorg' (pid 726)
NVRM: has the version 580.159.04, but this kernel module has
NVRM: the version 580.159.03. Please make sure that this
NVRM: kernel module and all NVIDIA driver components
NVRM: have the same version.
Jun 01 19:32:49 theep kernel: mdio_devres mxm_wmi i2c_i801 mei_me snd pmt_telemetry libphy i2c_smbus spi_intel_pci pmt_discovery soundcore mei i2c_mux spi_intel intel_pch_thermal mdio_bus pmt_class intel_pmc_ssram_telemetry intel_vsec>
Jun 01 19:32:54 theep kernel: NVRM: API mismatch: the client 'kwin_wayland' (pid 888)
NVRM: has the version 580.159.04, but this kernel module has
NVRM: the version 580.159.03. Please make sure that this
NVRM: kernel module and all NVIDIA driver components
NVRM: have the same version.
Jun 01 19:32:54 theep kernel: NVRM: API mismatch: the client 'Xwayland' (pid 964)
NVRM: has the version 580.159.04, but this kernel module has
NVRM: the version 580.159.03. Please make sure that this
NVRM: kernel module and all NVIDIA driver components
NVRM: have the same version.
Those last few errors continue on for a while and are actually something I found when I got back about an hour ago. dkms status gives this:
Failed to initialize NVML: Driver/library version mismatch
NVML library version: 580.159
I tried downgrading the packages to version 580.159.03 but I’m too much of a linux noob to know how to do that .-. and in any case it doesn’t really feel like the right solution…
I fixed the issue by downgrading the nvidia-580xx-dkms package to match the driver version found in /proc/driver/nvidia/version. I struggled for a while basically learning how pacman works but eventually I found that I could downgrade nvidia-580xx-utils first with the -d flag to skip dependency checks and then downgrade nvidia-580xx-dkms. After doing all this it works perfectly
