could not find module by name='off'

[graingert]

When I try to load the nvidia module it doesn't load.

$ sudo modprobe nvidia
modprobe: ERROR: ../libkmod/libkmod-module.c:832 kmod_module_insert_module() could not find module by name='off'
modprobe: ERROR: could not insert 'off': Unknown symbol in module, or unknown parameter (see dmesg)
$ sudo prime-select query
intel
$ sudo prime-offload
$ cat /var/log/prime-supported.log
Requires offloading
$ cat /var/log/prime-offload.log
Sorry the nvidia kernel module is not is loaded

ProblemType: Bug
DistroRelease: Ubuntu 18.04
Package: nvidia-kernel-common-390 390.48-0ubuntu3
ProcVersionSignature: Ubuntu 4.15.0-22.24-generic 4.15.17
Uname: Linux 4.15.0-22-generic x86_64
ApportVersion: 2.20.9-0ubuntu7.1
Architecture: amd64
CurrentDesktop: ubuntu:GNOME
Date: Thu May 31 09:42:42 2018
Dependencies:

InstallationDate: Installed on 2018-05-30 (0 days ago)
InstallationMedia: Ubuntu 18.04 LTS "Bionic Beaver" - Release amd64 (20180426)
ProcEnviron:
 TERM=xterm-256color
 PATH=(custom, no user)
 XDG_RUNTIME_DIR=<set>
 LANG=en_GB.UTF-8
 SHELL=/bin/bash
SourcePackage: nvidia-graphics-drivers-390
UpgradeStatus: No upgrade log present (probably fresh install)

[graingert (tagrain)]

Similar to https://bugs.launchpad.net/ubuntu/+source/bumblebee/+bug/1438106 but I'm not using bumblebee

The file in this nvidia-graphics-drivers-390 package still has those aliases

$ cat /etc/modprobe.d/nvidia-graphics-drivers.conf
blacklist nouveau
blacklist lbm-nouveau
alias nouveau off
alias lbm-nouveau off

[Launchpad Janitor (janitor)]

Status changed to 'Confirmed' because the bug affects multiple users.

[Twisted Lucidity (lucidlytwisted)]

Kubuntu 18.04 on a T430 with an nvidia 5400M.
Same issue with nvidia 390 and 396.
Not using Bumblebee.

[Zhou (fxiong)]

Ubuntu 18.04 On Precision 5530 Quadro P1000
Confirm the issue with nvidai 396.

I solve it by using Rocko (rockorequin)'s solution
Commenting out lines in /etc/modprobe.d/<nvida related>.conf fixed it:

alias <parameter> off

Original post: https://bugs.launchpad.net/ubuntu/+source/bumblebee/+bug/1438106

[Dan Darie (z-dan-3)]

Ubuntu 18.04 on Asus Zenbook Pro UX901V, GeForce GTX 960M, both nvidia 390 and 396.

Commenting the alias lines didn't work for me.

[Dan Darie (z-dan-3)]

Fixed it by commenting lines in 159-161 in /usr/bin/prime-select:

#alias nvidia off
#alias nvidia-drm off
#alias nvidia-modeset off'''

Also, comment lines 3-4 in /lib/modprobe.d/nvidia-graphics-drivers.conf

#alias nouveau off
#alias lbm-nouveau off

[Levente (levente12)]

Same problem with nvidia 410

[Igor Mineev (objatiegroba)]

Same problem with nvidia 410.
Reinstall not working.

[Igor Mineev (objatiegroba)]

Comment lines at /etc/modprobe.d/nvidia-...conf
not working
~$ sudo modprobe nvidia
same error:
modprobe: ERROR: ../libkmod/libkmod-module.c:832 kmod_module_insert_module() could not find module by name='off'
modprobe: ERROR: could not insert 'off': Unknown symbol in module, or unknown parameter (see dmesg)

[Vladislav Dalechin (vladik2738)]

I can confirm that I also have that bug with nvidia 410 and bumblebee.
Bumblebee tries to modprobe nvidia and fails with same error

[Vladislav Dalechin (vladik2738)]

I see only few solutions to this bug:
 - Somehow find file which has 'alias nvidia off', because I can't find any
 - Debug this kernel module (which is actually a very long process)

[Vladislav Dalechin (vladik2738)]

So i actually found one in /lib/modprobe.d/blacklist-nvidia.conf
Blacklisting it actually solved the bug, but my ubuntu is crashing after that, but that's another story.

[Matthias (maflobra)]

Same problem here, with Ubuntu 18.04, Nvidia drivers 410.
After commenting the following lines:

#alias nvidia off
#alias nvidia-drm off
#alias nvidia-modeset off

either in /usr/bin/prime-select (as suggested above) or directly in /lib/modprobe.d/blacklist-nvidia.conf I can load the driver, however, that also leads to crashing my system (as described by Vladislav Dalechin.

Any news on this?

[David Cimadevilla (davidcim)]

Just upgraded from Ubuntu 16 to 18.04.

In my case the problem arises when I unload the driver and then try to load it again. After the error dmesg states the possible cause:

[ 966.030549] NVRM: API mismatch: the client has the version 390.116, but
               NVRM: this kernel module has the version 384.130. Please
               NVRM: make sure that this kernel module and all NVIDIA driver
               NVRM: components have the same version.

[Thomas Ward (teward)]

Similar activity on 440 version of drivers.

[Victor Saase (vsaase)]

Had the same problem, the reason was that prime-select was set to intel instead of nvidia.