Arch Linux

Please read this before reporting a bug:
https://wiki.archlinux.org/title/Bug_reporting_guidelines

Do NOT report bugs when a package is just outdated, or it is in the AUR. Use the 'flag out of date' link on the package page, or the Mailing List.

REPEAT: Do NOT report bugs for outdated packages!
Tasklist

FS#77455 - Updating nvidia-utils & nvidia-dkms breaks Vulkan support and causes screen glitches.

Attached to Project: Arch Linux
Opened by Patrick Lineruth (WaCrex) - Friday, 10 February 2023, 14:42 GMT
Last edited by Toolybird (Toolybird) - Friday, 10 February 2023, 20:44 GMT
Task Type Bug Report
Category Packages: Extra
Status Closed
Assigned To No-one
Architecture x86_64
Severity Medium
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

I was too tired to update my system yesterday before turning the system off. But today when I installed the updates, it completely broke everything.

* Vulkan stopped working (can't run vkcube mentioned something about ICD in the console when I tried).
* Chromium browser wouldn't start, I'm using chromium-wayland-vaapi (AUR) with this chromium-flags.conf:
--enable-features=WebUIDarkMode
--enable-features=VaapiVideoEncoder,VaapiVideoDecoder,CanvasOopRasterization,Vulkan,UseSkiaRenderer,UseOzonePlatform
--enable-zero-copy
--use-gl=egl
--ignore-gpu-blocklist
--enable-oop-rasterization
--enable-raw-draw
--enable-gpu-rasterization
--use-vulkan
--disable-reading-from-canvas
--disable-sync-preferences
--ozone-platform=wayland
* My external screen was only half rendered when signing in on gdm (the other half of the screen was black).
* The desktop felt sluggish and glitchy.

So I did a snapshot recovery in Timeshift to my automatic snapshot before the update, and everything worked fine again.
These are the packages that were updated before I started having issues.

Package (12) Old Version New Version Net Change

multilib/lib32-nvidia-utils 525.85.05-1 525.89.02-1 0.00 MiB
multilib/lib32-opencl-nvidia 525.85.05-1 525.89.02-1 0.00 MiB
multilib/lib32-sdl2 2.26.2-1 2.26.3-1 0.00 MiB
multilib/lib32-sdl2_image 2.6.2-2 2.6.3-1 0.00 MiB
multilib/lib32-sdl2_mixer 2.6.2-1 2.6.3-1 0.00 MiB
extra/libxfce4ui 4.18.1-1 4.18.2-1 0.08 MiB
extra/libxnvctrl 525.85.05-1 525.89.02-1 0.00 MiB
extra/lmdb 0.9.29-1 0.9.30-1 -0.01 MiB
extra/nvidia-dkms 525.85.05-3 525.89.02-1 -0.03 MiB
extra/nvidia-settings 525.85.05-1 525.89.02-1 0.00 MiB
extra/nvidia-utils 525.85.05-3 525.89.02-1 0.01 MiB
This task depends upon

Closed by  Toolybird (Toolybird)
Friday, 10 February 2023, 20:44 GMT
Reason for closing:  Not a bug
Additional comments about closing:  Reporter says "Problem solved, turned out I had forgotten to run "sudo mkinitcpio -P" after GPU driver update."
Comment by Doug Newgard (Scimmia) - Friday, 10 February 2023, 14:55 GMT
Did you reboot? nvidia-utils needs to match the module that's loaded, not the one that's installed.
Comment by Patrick Lineruth (WaCrex) - Friday, 10 February 2023, 17:18 GMT
Yes, I rebooted right after installing the update, I always want to test with a reboot after things like GPU driver updates or similar.

----------

Also, I forgot to mention this in the post:

[wacrex@LEGION ~]$ uname -a
Linux LEGION 6.1.10-zen1-1-zen #1 ZEN SMP PREEMPT_DYNAMIC Mon, 06 Feb 2023 09:28:06 +0000 x86_64 GNU/Linux

[wacrex@LEGION ~]$ lspci | grep VGA
01:00.0 VGA compatible controller: NVIDIA Corporation GA104M [GeForce RTX 3070 Mobile / Max-Q] (rev a1)
Comment by Patrick Lineruth (WaCrex) - Friday, 10 February 2023, 17:36 GMT
Also, I just did some digging in journalctl between the time the timeshift-autosnap snapshot was taken (2023-02-10 15:01:27) and me restoring the system to the snapshot (2023-02-10 15:08:11):

[wacrex@LEGION ~]$ journalctl -p err
...
Feb 10 15:03:24 LEGION systemd-coredump[53709]: Failed to connect to coredump service: Connection refused
Feb 10 15:03:24 LEGION gdm[727]: Gdm: Failed to contact accountsservice: Error calling StartServiceByName for org.freedesktop.Accounts: Refusing activation, D-Bus is shutting down.
-- Boot b8797282b65648daaa2e6922baa10bd4 --
Feb 10 15:05:19 archlinux kernel: ACPI BIOS Error (bug): Could not resolve symbol [\_SB.PCI0.PB2], AE_NOT_FOUND (20220331/dswload2-162)
Feb 10 15:05:19 archlinux kernel: ACPI Error: AE_NOT_FOUND, During name lookup/catalog (20220331/psobject-220)
Feb 10 15:05:20 archlinux kernel:
Feb 10 15:05:23 LEGION kernel: hid-generic 0003:0D8C:0006.0007: No inputs registered, leaving
Feb 10 15:05:23 LEGION kernel: rtw89_8852ae 0000:04:00.0: failed to early request firmware: -2
Feb 10 15:05:24 LEGION kernel: Bluetooth: hci0: Failed to read codec capabilities (-22)
Feb 10 15:05:24 LEGION bluetoothd[703]: src/plugin.c:plugin_init() Failed to init vcp plugin
Feb 10 15:05:24 LEGION bluetoothd[703]: src/plugin.c:plugin_init() Failed to init mcp plugin
Feb 10 15:05:24 LEGION bluetoothd[703]: src/plugin.c:plugin_init() Failed to init bap plugin
Feb 10 15:05:24 LEGION bluetoothd[703]: Failed to set mode: Failed (0x03)
Feb 10 15:05:25 LEGION gnome-session-binary[802]: Unrecoverable failure in required component org.gnome.Shell.desktop
Feb 10 15:05:25 LEGION gdm-launch-environment][765]: GLib-GObject: g_object_unref: assertion 'G_IS_OBJECT (object)' failed
Feb 10 15:05:43 LEGION gdm-password][1410]: gkr-pam: unable to locate daemon control file
Feb 10 15:05:47 LEGION systemd[1424]: Failed to start Application launched by gnome-session-binary.
Feb 10 15:05:47 LEGION systemd[1424]: Failed to start Application launched by gnome-session-binary.
Feb 10 15:05:47 LEGION systemd[1424]: Failed to start Application launched by gnome-session-binary.
Feb 10 15:05:49 LEGION systemd[1424]: Failed to start Application launched by gnome-session-binary.
Feb 10 15:05:51 LEGION gdm-launch-environment][883]: GLib-GObject: g_object_unref: assertion 'G_IS_OBJECT (object)' failed
Feb 10 15:05:55 LEGION kernel: [drm:nv_drm_connector_get_modes [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to get mode at modeIndex 1 of NvKmsKapiDisplay 0x00001000
Feb 10 15:05:55 LEGION kernel: [drm:nv_drm_connector_get_modes [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to get mode at modeIndex 3 of NvKmsKapiDisplay 0x00002000
Feb 10 15:06:52 LEGION sudo[4265]: wacrex : 1 incorrect password attempt ; TTY=pts/0 ; PWD=/home/WaCrex ; USER=root ; COMMAND=/usr/bin/pacman -S firefox
Feb 10 15:08:43 LEGION systemd-coredump[5204]: Failed to connect to coredump service: Connection refused
Feb 10 15:08:43 LEGION gdm[737]: Gdm: Failed to contact accountsservice: Error calling StartServiceByName for org.freedesktop.Accounts: Refusing activation, D-Bus is shutting down.
Feb 10 15:08:44 LEGION pipewire[1714]: spa.alsa: 'hdmi:0': playback open failed: No such file or directory
-- Boot be568c0919d54f64b62bdfdfc3019916 --
...

[wacrex@LEGION ~]$ journalctl | grep nvidia
...
Feb 10 15:04:54 LEGION systemd[1]: Stopping Load/Save Screen Backlight Brightness of backlight:nvidia_0...
Feb 10 15:04:54 LEGION systemd[1]: systemd-backlight@backlight:nvidia_0.service: Deactivated successfully.
Feb 10 15:04:54 LEGION systemd[1]: Stopped Load/Save Screen Backlight Brightness of backlight:nvidia_0.
Feb 10 15:04:54 LEGION audit[1]: SERVICE_STOP pid=1 uid=0 auid=4294967295 ses=4294967295 msg='unit=systemd-backlight@backlight:nvidia_0 comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Feb 10 15:04:54 LEGION kernel: audit: type=1131 audit(1676037894.896:583): pid=1 uid=0 auid=4294967295 ses=4294967295 msg='unit=systemd-backlight@backlight:nvidia_0 comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Feb 10 15:05:19 archlinux kernel: Command line: BOOT_IMAGE=/vmlinuz-linux-zen root=UUID=8c0b630c-693a-45a2-8278-cbbea4da6c9e rw rootflags=subvol=@ resume=UUID=1d4dede0-d618-477a-80c2-3db41eba46c6 nvidia_drm.modeset=1 loglevel=3 quiet
Feb 10 15:05:19 archlinux kernel: Kernel command line: BOOT_IMAGE=/vmlinuz-linux-zen root=UUID=8c0b630c-693a-45a2-8278-cbbea4da6c9e rw rootflags=subvol=@ resume=UUID=1d4dede0-d618-477a-80c2-3db41eba46c6 nvidia_drm.modeset=1 loglevel=3 quiet
Feb 10 15:05:20 archlinux kernel: nvidia: loading out-of-tree module taints kernel.
Feb 10 15:05:20 archlinux kernel: nvidia: module license 'NVIDIA' taints kernel.
Feb 10 15:05:20 archlinux kernel: nvidia: module verification failed: signature and/or required key missing - tainting kernel
Feb 10 15:05:20 archlinux kernel: nvidia-nvlink: Nvlink Core is being initialized, major device number 235
Feb 10 15:05:20 archlinux kernel: nvidia 0000:01:00.0: vgaarb: changed VGA decodes: olddecodes=io+mem,decodes=none:owns=none
Feb 10 15:05:20 archlinux systemd-modules-load[232]: Inserted module 'nvidia'
Feb 10 15:05:20 archlinux systemd-modules-load[232]: Inserted module 'nvidia_modeset'
Feb 10 15:05:20 archlinux kernel: nvidia-modeset: Loading NVIDIA Kernel Mode Setting Driver for UNIX platforms 525.85.05 Sat Jan 14 00:40:03 UTC 2023
Feb 10 15:05:20 archlinux kernel: [drm] [nvidia-drm] [GPU ID 0x00000100] Loading driver
Feb 10 15:05:20 archlinux kernel: nvidia_uvm: module uses symbols nvUvmInterfaceDisableAccessCntr from proprietary module nvidia, inheriting taint.
Feb 10 15:05:21 archlinux kernel: nvidia-uvm: Loaded the UVM driver, major device number 511.
Feb 10 15:05:21 archlinux systemd-modules-load[232]: Inserted module 'nvidia_uvm'
Feb 10 15:05:21 archlinux systemd-modules-load[232]: Inserted module 'nvidia_drm'
Feb 10 15:05:21 archlinux kernel: [drm] Initialized nvidia-drm 0.0.0 20160202 for 0000:01:00.0 on minor 0
Feb 10 15:05:23 LEGION systemd-udevd[466]: nvidia: Process '/usr/bin/bash -c '/usr/bin/mknod -Z -m 666 /dev/nvidiactl c $(grep nvidia-frontend /proc/devices | cut -d \ -f 1) 255'' failed with exit code 1.
Feb 10 15:05:23 LEGION systemd-udevd[466]: nvidia: Process '/usr/bin/bash -c 'for i in $(cat /proc/driver/nvidia/gpus/*/information | grep Minor | cut -d \ -f 4); do /usr/bin/mknod -Z -m 666 /dev/nvidia${i} c $(grep nvidia-frontend /proc/devices | cut -d \ -f 1) ${i}; done'' failed with exit code 1.
Feb 10 15:05:23 LEGION systemd[1]: Starting Load/Save Screen Backlight Brightness of backlight:nvidia_0...
Feb 10 15:05:23 LEGION systemd[1]: Finished Load/Save Screen Backlight Brightness of backlight:nvidia_0.
Feb 10 15:05:23 LEGION audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295 ses=4294967295 msg='unit=systemd-backlight@backlight:nvidia_0 comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Feb 10 15:05:25 LEGION gnome-shell[826]: Added device '/dev/dri/card0' (nvidia-drm) using atomic mode setting.
Feb 10 15:05:25 LEGION /usr/lib/gdm-x-session[888]: Kernel command line: BOOT_IMAGE=/vmlinuz-linux-zen root=UUID=8c0b630c-693a-45a2-8278-cbbea4da6c9e rw rootflags=subvol=@ resume=UUID=1d4dede0-d618-477a-80c2-3db41eba46c6 nvidia_drm.modeset=1 loglevel=3 quiet
Feb 10 15:05:25 LEGION /usr/lib/gdm-x-session[888]: (**) OutputClass "nvidia" ModulePath extended to "/usr/lib/nvidia/xorg,/usr/lib/xorg/modules,/usr/lib/xorg/modules"
Feb 10 15:05:25 LEGION /usr/lib/gdm-x-session[888]: (II) Applying OutputClass "nvidia" to /dev/dri/card0
Feb 10 15:05:25 LEGION /usr/lib/gdm-x-session[888]: loading driver: nvidia
Feb 10 15:05:25 LEGION /usr/lib/gdm-x-session[888]: (==) Matched nvidia as autoconfigured driver 0
Feb 10 15:05:25 LEGION /usr/lib/gdm-x-session[888]: (II) LoadModule: "nvidia"
Feb 10 15:05:25 LEGION /usr/lib/gdm-x-session[888]: (II) Loading /usr/lib/xorg/modules/drivers/nvidia_drv.so
Feb 10 15:05:25 LEGION /usr/lib/gdm-x-session[888]: (II) Module nvidia: vendor="NVIDIA Corporation"
Feb 10 15:05:25 LEGION /usr/lib/gdm-x-session[888]: (II) Applying OutputClass "nvidia" options to /dev/dri/card0
Feb 10 15:05:45 LEGION /usr/lib/gdm-x-session[1452]: Kernel command line: BOOT_IMAGE=/vmlinuz-linux-zen root=UUID=8c0b630c-693a-45a2-8278-cbbea4da6c9e rw rootflags=subvol=@ resume=UUID=1d4dede0-d618-477a-80c2-3db41eba46c6 nvidia_drm.modeset=1 loglevel=3 quiet
Feb 10 15:05:45 LEGION /usr/lib/gdm-x-session[1452]: (**) OutputClass "nvidia" ModulePath extended to "/usr/lib/nvidia/xorg,/usr/lib/xorg/modules,/usr/lib/xorg/modules"
Feb 10 15:05:45 LEGION /usr/lib/gdm-x-session[1452]: (II) Applying OutputClass "nvidia" to /dev/dri/card0
Feb 10 15:05:45 LEGION /usr/lib/gdm-x-session[1452]: loading driver: nvidia
Feb 10 15:05:45 LEGION /usr/lib/gdm-x-session[1452]: (==) Matched nvidia as autoconfigured driver 0
Feb 10 15:05:45 LEGION /usr/lib/gdm-x-session[1452]: (II) LoadModule: "nvidia"
Feb 10 15:05:45 LEGION /usr/lib/gdm-x-session[1452]: (II) Loading /usr/lib/xorg/modules/drivers/nvidia_drv.so
Feb 10 15:05:45 LEGION /usr/lib/gdm-x-session[1452]: (II) Module nvidia: vendor="NVIDIA Corporation"
Feb 10 15:05:45 LEGION /usr/lib/gdm-x-session[1452]: (II) Applying OutputClass "nvidia" options to /dev/dri/card0
Feb 10 15:05:55 LEGION kernel: [drm:nv_drm_connector_get_modes [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to get mode at modeIndex 1 of NvKmsKapiDisplay 0x00001000
Feb 10 15:05:55 LEGION kernel: [drm:nv_drm_connector_get_modes [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to get mode at modeIndex 3 of NvKmsKapiDisplay 0x00002000
...
Comment by Doug Newgard (Scimmia) - Friday, 10 February 2023, 17:50 GMT
Feb 10 15:05:20 archlinux kernel: nvidia-modeset: Loading NVIDIA Kernel Mode Setting Driver for UNIX platforms 525.85.05 Sat Jan 14 00:40:03 UTC 2023

You're saying this was before you restored the snapshot? Notice the version. Are you using early KMS (kms hook in mkinitcpio, or specific modules)? If so, did you rebuild the initramfs after updating nvidia?
Comment by Patrick Lineruth (WaCrex) - Friday, 10 February 2023, 18:24 GMT
Thanks a LOT for the comment, that seems to have solved it for me.. I re-did the install and did an sudo mkinitcpio -P afterwards and rebooted the system.. it's now working just fine with the update... I seem to have forgotten to add a pacman hook for running mkinitcpio -P after my last re-install a few days ago^^

Everything is working fine now :)
Comment by Patrick Lineruth (WaCrex) - Friday, 10 February 2023, 18:27 GMT
It's a bit embarrassing that it was just a simple "sudo mkinitcpio -P" that was missing ^^

Once again, thanks :)

Loading...