dkms false-positive error, since 37740, installing kernel module the first time

[marioroy]

I met to submit an issue at the time and remembered today.

37740 change log:

dkms  3.0.8-22  ->  3.0.9-23

I now pass the --force argument to dkms to rid of the false-positive. For the NVIDIA installer, I now output standard-error output to /dev/null to rid of the false-positive. dkms worked wonderfully before.

-> Would you like to register the kernel module sources with DKMS? This will allow DKMS to automatically build a new module, if your kernel changes later. (Answer: Yes)
-> Registering the kernel modules with DKMS:
ERROR: Failed to run `/usr/bin/dkms install --no-depmod -m nvidia -v 520.61.05 -k 5.15.84-27.lts2021`:
nvidia.ko:
Running module version sanity check.
Module version 520.61.05 for nvidia.ko
exactly matches what is already found in kernel 5.15.84-27.lts2021.
DKMS will not replace this module.
You may override by specifying --force.
   
nvidia-uvm.ko:
Running module version sanity check.
Module version 520.61.05 for nvidia-uvm.ko
exactly matches what is already found in kernel 5.15.84-27.lts2021.
DKMS will not replace this module.
You may override by specifying --force.
   
nvidia-modeset.ko:
Running module version sanity check.
Module version 520.61.05 for nvidia-modeset.ko
exactly matches what is already found in kernel 5.15.84-27.lts2021.
DKMS will not replace this module.
You may override by specifying --force.
    
nvidia-drm.ko:
Running module version sanity check.
Module version 520.61.05 for nvidia-drm.ko
exactly matches what is already found in kernel 5.15.84-27.lts2021.
DKMS will not replace this module.
You may override by specifying --force.
    
nvidia-peermem.ko:
Running module version sanity check.
Module version 520.61.05 for nvidia-peermem.ko
exactly matches what is already found in kernel 5.15.84-27.lts2021.
DKMS will not replace this module.
You may override by specifying --force.
Error! Installation aborted.
-> Error.
WARNING: Failed to register the NVIDIA kernel modules with DKMS. The NVIDIA kernel modules were installed, but will not be automatically rebuilt if you change your kernel.

Fortunately, the NVIDIA installer continues to completion and exits with a status of 0. The automation script subsequently runs dkms with the --force argument.

[marioroy]

The /usr/bin/dkms-new-kernel.sh script, without the --force argument, seems to be working when a new kernel is installed; e.g. the system boots into the graphical display using NVIDIA driver. The false-positive occurs the first time installing the kernel module.

[marioroy]

This too, is a false-positive in the NVIDIA installer log.

WARNING: Failed to register the NVIDIA kernel modules with DKMS. The NVIDIA kernel modules were installed, but will not be automatically rebuilt if you change your kernel.

The NVIDIA modules are re-built during swupd update when a new kernel is installed. No issues with /usr/bin/dkms-new-kernel.sh.

[marioroy]

The issue persists in CL 37850. The NVIDIA driver automation succeeds regardless, described above.

Blessings and Grace

[cr0wg4n]

fedora 6.1.12-200.fc37, same issue

But the NVIDIA driver is installed anyway

[ddsmith2-eprod]

I have the same issue except there is no false positive. After a kernel update the driver is no longer available.

DKMS shows its working:

$ dkms status
nvidia/525.85.12, 4.18.0-425.3.1.el8.x86_64, x86_64: built

The Nvidia driver has disappeared:

$ lsmod |grep nv
nvmet_fc 40960 1 lpfc
nvmet 114688 1 nvmet_fc
nvme_fc 53248 1 lpfc
nvme_fabrics 24576 1 nvme_fc
nvme_core 139264 2 nvme_fc,nvme_fabrics
t10_pi 16384 3 nvmet,sd_mod,nvme_core

DKMS's status above is pointing to the previous kernel:

$ uname -a
Linux pdds2wk07corp.redacted.com 4.18.0-425.19.2.el8_7.x86_64 #1 SMP Fri Mar 17 01:52:38 EDT 2023 x86_64 x86_64 x86_64 GNU/Linux

[fenrus75]

Something is VERY wrong in your report there since we havent had a kernel version starting with a 4 in a long time, and even kernel versions with a 5 are well behind us (except for the LTS-1 release) Are you sure the clr-boot-manager agent is working correctly?

…

On Wed, Apr 19, 2023 at 5:31 AM ddsmith2-eprod ***@***.***> wrote: I have the same issue except there is no false positive. After a kernel update the driver is no longer available. DKMS shows its working: $ dkms status nvidia/525.85.12, 4.18.0-425.3.1.el8.x86_64, x86_64: built The Nvidia driver has disappeared: $ lsmod |grep nv nvmet_fc 40960 1 lpfc nvmet 114688 1 nvmet_fc nvme_fc 53248 1 lpfc nvme_fabrics 24576 1 nvme_fc nvme_core 139264 2 nvme_fc,nvme_fabrics t10_pi 16384 3 nvmet,sd_mod,nvme_core — Reply to this email directly, view it on GitHub <#2792 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAJ54FI5ABZYK4RDJPGDZGLXB7LJPANCNFSM6AAAAAATGDAUCE> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[ddsmith2-eprod]

My bad, you are correct. I found this post with Google and thought this issue tracker was for dkms. My OS is RHEL so you can just ignore my response.