Issues with Stable Update 2025.05.14

Support System Updates
1

Heyo! I’m having some issues with the latest update. I haven’t had issues with previous updates, so it’s a bit baffling.

(I didn’t realize I replied to a specific post, that was not intentional)

I have also tried updating my pacman mirrors.

Please let me know what info I could provide if I missed anything useful, this is my first time posting in a forum.

My hardware:

  • AMD 5800x
  • EVGA Nvidia 3070ti

I did try updating video-nvidia with mhwd, but there were conflicts, and I foolishly forgot to grab the logs of what they were. I assume this is due to me trying to use the repo below.

MY guess is that it’s a combination of gcc15 updating and how I currently install Nvidia drivers.

I’ve been using this repo as an easy way to update my nvidia drivers.

However, I know this commit exist for the GCC 15 version

But, when I try rebuilding these AFTER the update inside TTY, I only get errors, it’s possible I’m somehow failing to update properly, but it seems hard to type makepkg -si (and I updated the code with latest)

Updating and getting stuck

The general flow of this went as such:

  • Reboot after update as per usual
  • Don’t see booting animation like I usually do.
  • Open TTY2
  • Check journalctl logs, don’t see anything out of the ordinary.
  • Reboot
  • Check journalctl logs, notice it says Nvidia kernel modules can no longer be found.
  • Use timeshift to restore to point just before update.

Here’s my driver info before updating. Note here, these did not change after running the update.

~> mhwd-kernel -li
Currently running: 6.12.25-1-MANJARO (linux612)
The following kernels are installed in your system:
   * linux54
   * linux612
~> 

~> mhwd -l
> 0000:09:00.0 (0300:10de:2482) Display controller nVidia Corporation:
--------------------------------------------------------------------------------
                  NAME               VERSION          FREEDRIVER           TYPE
--------------------------------------------------------------------------------
          video-nvidia            2025.01.13               false            PCI
    video-nvidia-470xx            2023.03.23               false            PCI
           video-linux            2024.05.06                true            PCI
     video-modesetting            2020.01.13                true            PCI
            video-vesa            2017.03.12                true            PCI


> 0000:05:00.0 (0200:10ec:8168) Network controller Realtek Semiconductor Co., Ltd.:
--------------------------------------------------------------------------------
                  NAME               VERSION          FREEDRIVER           TYPE
--------------------------------------------------------------------------------
         network-r8168            2023.09.12                true            PCI

~>

Here are my journalctl logs after the first reboot. To me, these aren’t anything crazy or to suspect issues from.

May 14 19:09:04 kraken (sd-worker)[40301]: Failed start worker process: No such file or directory
May 14 19:09:04 kraken (sd-worker)[40302]: Failed start worker process: Text file busy
May 14 19:11:58 kraken dbus-broker-launch[813]: Activation request for 'org.bluez' failed.
May 14 19:11:58 kraken dbus-broker-launch[813]: Activation request for 'org.freedesktop.UDisks2' failed.
May 14 19:11:58 kraken dbus-broker-launch[813]: Activation request for 'org.bluez' failed.
May 14 19:11:58 kraken systemd-coredump[99970]: Failed to send pidfd: Connection reset by peer
May 14 19:11:58 kraken dbus-broker-launch[813]: Activation request for 'org.freedesktop.UPower' failed.
May 14 19:11:59 kraken dbus-broker-launch[813]: Activation request for 'org.freedesktop.nm_dispatcher' failed.
May 14 19:11:59 kraken dbus-broker-launch[813]: Activation request for 'org.bluez' failed.
May 14 19:11:59 kraken systemd-coredump[100227]: Failed to send coredump datagram: Connection reset by peer
May 14 19:11:59 kraken systemd-coredump[100318]: Failed to send coredump fd: Broken pipe
May 14 19:12:00 kraken (sd-pam)[1015]: pam_systemd(systemd-user:session): Failed to release session: No session '2' known
May 14 19:12:00 kraken dbus-broker-launch[813]: Activation request for 'org.freedesktop.nm_dispatcher' failed.
May 14 19:12:00 kraken dbus-broker-launch[813]: Activation request for 'org.freedesktop.nm_dispatcher' failed.
May 14 19:12:01 kraken kernel: watchdog: watchdog0: watchdog did not stop!

After this I rebooted. Which I was met with the same screen where I see my motherboards logo.

Checking logs again, they changed to this.

May 14 19:12:26 kraken systemd-modules-load[412]: Failed to find module 'nvidia'
May 14 19:12:26 kraken systemd-modules-load[412]: Failed to find module 'nvidia_drm'
May 14 19:12:26 kraken systemd-modules-load[412]: Failed to find module 'nvidia-uvm'
May 14 19:23:22 kraken dbus-broker-launch[817]: Activation request for 'org.freedesktop.nm_dispatcher' failed.

And here are some pacman logs.

[2025-05-14T19:10:04-0500] [ALPM-SCRIPTLET] ==> dkms install --no-depmod nvidia/570.123.14 -k 5.4.293-1-MANJARO
[2025-05-14T19:10:24-0500] [ALPM-SCRIPTLET] 
[2025-05-14T19:10:24-0500] [ALPM-SCRIPTLET] Error! Bad return status for module build on kernel: 5.4.293-1-MANJARO (x86_64)
[2025-05-14T19:10:24-0500] [ALPM-SCRIPTLET] Consult /var/lib/dkms/nvidia/570.123.14/build/make.log for more information.
[2025-05-14T19:10:24-0500] [ALPM-SCRIPTLET] ==> WARNING: `dkms install --no-depmod nvidia/570.123.14 -k 5.4.293-1-MANJARO' exited 10
[2025-05-14T19:10:24-0500] [ALPM-SCRIPTLET] ==> dkms install --no-depmod v4l2loopback/0.14.0 -k 6.12.28-1-MANJARO
[2025-05-14T19:10:25-0500] [ALPM-SCRIPTLET] Deprecated feature: CLEAN (/usr/src/v4l2loopback-0.14.0/dkms.conf)
[2025-05-14T19:10:27-0500] [ALPM-SCRIPTLET] ==> dkms install --no-depmod v4l2loopback/0.14.0 -k 5.4.293-1-MANJARO
[2025-05-14T19:10:28-0500] [ALPM-SCRIPTLET] Deprecated feature: CLEAN (/var/lib/dkms/v4l2loopback/0.14.0/source/dkms.conf)
[2025-05-14T19:10:30-0500] [ALPM-SCRIPTLET] ==> dkms install --no-depmod nvidia/570.123.14 -k 6.12.28-1-MANJARO
[2025-05-14T19:10:49-0500] [ALPM-SCRIPTLET] 
[2025-05-14T19:10:49-0500] [ALPM-SCRIPTLET] Error! Bad return status for module build on kernel: 6.12.28-1-MANJARO (x86_64)
[2025-05-14T19:10:49-0500] [ALPM-SCRIPTLET] Consult /var/lib/dkms/nvidia/570.123.14/build/make.log for more information.
[2025-05-14T19:10:49-0500] [ALPM-SCRIPTLET] ==> WARNING: `dkms install --no-depmod nvidia/570.123.14 -k 6.12.28-1-MANJARO' exited 10
3

did you check that on your system? GCC15 is not nice to Nvidia drivers. However if you use mhwd it normally should install the proper driver for you as we had packaged it. DKMS can be a hit or miss. However 570 series we use dkms to build our matching kernel drivers.

Check what parts you have installed on your system via pacman -Q | grep nvidia. Avoid 3rd party PKGBUILDs unless you know what you do.

You also find our driver sources here: Packages / extra / nvidia-utils · GitLab

1 Like
4

Ah, I forgot to post that log, but I think you’re right about the fact it doesn’t play nice with gcc15.

CC [M]  /var/lib/dkms/nvidia/570.123.14/build/nvidia-drm/nvidia-drm-gem-user-memory.o
/var/lib/dkms/nvidia/570.123.14/build/nvidia-drm/nvidia-drm-connector.c: In function ‘__nv_drm_connector_detect_internal’:
/var/lib/dkms/nvidia/570.123.14/build/nvidia-drm/nvidia-drm-connector.c:181:66: error: macro "drm_connector_for_each_possible_encoder" requires 3 arguments, but only 2 given
  181 |     nv_drm_connector_for_each_possible_encoder(connector, encoder) {
      |                                                                  ^
In file included from ./include/drm/drm_modes.h:33,
                 from ./include/drm/drm_crtc.h:40,
                 from ./include/drm/drmP.h:68,
                 from /var/lib/dkms/nvidia/570.123.14/build/nvidia-drm/nvidia-drm-helper.h:31,
                 from /var/lib/dkms/nvidia/570.123.14/build/nvidia-drm/nvidia-drm-connector.c:27:
./include/drm/drm_connector.h:1613:9: note: macro "drm_connector_for_each_possible_encoder" defined here
 1613 | #define drm_connector_for_each_possible_encoder(connector, encoder, __i) \
      |         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/var/lib/dkms/nvidia/570.123.14/build/nvidia-drm/nvidia-drm-helper.h:510:5: error: ‘drm_connector_for_each_possible_encoder’ undeclared (first use in this function); did you mean ‘drm_connector_has_possible_encoder’?
  510 |     drm_connector_for_each_possible_encoder(connector, encoder)
      |     ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/var/lib/dkms/nvidia/570.123.14/build/nvidia-drm/nvidia-drm-connector.c:181:5: note: in expansion of macro ‘nv_drm_connector_for_each_possible_encoder’
  181 |     nv_drm_connector_for_each_possible_encoder(connector, encoder) {
      |     ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/var/lib/dkms/nvidia/570.123.14/build/nvidia-drm/nvidia-drm-helper.h:510:5: note: each undeclared identifier is reported only once for each function it appears in
  510 |     drm_connector_for_each_possible_encoder(connector, encoder)
      |     ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/var/lib/dkms/nvidia/570.123.14/build/nvidia-drm/nvidia-drm-connector.c:181:5: note: in expansion of macro ‘nv_drm_connector_for_each_possible_encoder’
  181 |     nv_drm_connector_for_each_possible_encoder(connector, encoder) {
      |     ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
/var/lib/dkms/nvidia/570.123.14/build/nvidia-drm/nvidia-drm-connector.c:181:68: error: expected ‘;’ before ‘{’ token
  181 |     nv_drm_connector_for_each_possible_encoder(connector, encoder) {
      |                                                                    ^
/var/lib/dkms/nvidia/570.123.14/build/nvidia-drm/nvidia-drm-connector.c:163:25: warning: unused variable ‘encoder’ [-Wunused-variable]
  163 |     struct drm_encoder *encoder;
      |                         ^~~~~~~
make[2]: *** [scripts/Makefile.build:263: /var/lib/dkms/nvidia/570.123.14/build/nvidia-drm/nvidia-drm-connector.o] Error 1
make[2]: *** Waiting for unfinished jobs....
make[1]: *** [Makefile:1755: /var/lib/dkms/nvidia/570.123.14/build] Error 2
make: *** [Makefile:115: modules] Error 2

# exit code: 2
# elapsed time: 00:00:28

Here’s the output from that.

The thing I linked at the start is how I’ve been lazily handling nvidia drivers, I basically just looked for arch based solutions and rolled with it.

Here’s the output of pacman -Q | grep nvidia

~> pacman -Q | grep nvidia
lib32-nvidia-dev-utils-tkg 570.123.14-263
lib32-opencl-nvidia-dev-tkg 570.123.14-263
mhwd-nvidia 570.144-2
mhwd-nvidia-390xx 390.157-19
mhwd-nvidia-470xx 470.256.02-10
nvidia-dev-dkms-tkg 570.123.14-263
nvidia-dev-settings-tkg 570.123.14-263
nvidia-dev-utils-tkg 570.123.14-263
nvidia-driver-assistant 0.21.51.03-1
opencl-nvidia-dev-tkg 570.123.14-263

After doing that, it reminded me that these two were the ones causing conflicts when I tried using mhwd earlier

lib32-nvidia-dev-utils-tkg 570.123.14-263
lib32-opencl-nvidia-dev-tkg 570.123.14-263

So maybe I can remove those and try again.

5

So, even after uninstalling the dev packages, upgrading, still seeing the same behavior as described in the first post.

Doing the mhwd command just says it’s skipping.

~> sudo mhwd -a pci nonfree 0300
> Skipping already installed config 'video-nvidia' for device: 0000:09:00.0 (0300:10de:2482) Display controller nVidia Corporation GA104 [GeForce RTX 3070 Ti]
~>

It’s making me question if something else is the real issue.

It’s still concerning that it takes an additional reboot to see kernel module fails for nvidia / nvidia_drm.

6

Those dkms packages are from the AUR or another 3rd party. Our packages we provide are patched as needed… consult the tkg maintainer.