Fedora 37/36/35 NVIDIA Drivers Install Guide [530.30.02 / 525.89.02 / 520.56.06 / 515.86.01 / 510.108.03 / 470.161.03 / 390.157 / 340.108] - Comment Page: 9
This is guide, howto install NVIDIA proprietary drivers (manually using .run installer) on Fedora 37/36/35/34/33/32 and disable Nouveau driver. This guide works with GeForce 8/9/200/300/400/500/600/700/800/900/10/20/30/40 series cards.
GeForce RTX 40 series cards works with 530.xx, 525.xx, 520.xx NVIDIA drivers, (RTX 4090)
GeForce RTX 30 series cards works with 530.xx, 525.xx, 520.xx, 515.xx, 510.xx and 470.xx NVIDIA drivers, (RTX 3090, RTX 3080 and RTX 3070, RTX 3060, RTX 3060 Ti)
GeForce RTX 20 series cards works with 530.xx, 525.xx, 520.xx, 515.xx, 510.xx and 470.xx NVIDIA drivers (RTX 2080 Ti, RTX 2080, RTX 2070 Ti, RTX 2070, RTX 2060)
GeForce GT/GTX 600/700/800/900/10 series...
Thank you for the writeup. I got up to the part where I have to reboot in runlevel 3 (where I followed the instructions from http://www.if-not-true-then-false.com/2012/howto-change-runlevel-on-grub2/). I am stuck at a black screen with a blinking cursor. I tried to boot in safe mode with runlevel 3 and 1 but it is not working. Any ideas what may be wrong? Thanks.
I have an NVidia GeForce GTX 670 card. I installed 304xx akmod drivers and still was unable to get things up and running. As stated on other forums, switched to the 340xx akmod drivers and manually disabled the integrated graphics (using the jumper on the MB) and things work just fine with Gnome and Fedora 23 now.
For some reason get the following,would appreciate if you could help
Running F23 Gnome 3.18.3
GPU EVGA780Ti
So all should work
[[email protected] Downloads] # chmod +x /path/to/NVIDIA-Linux-*.run
chmod: cannot access ‘/path/to/NVIDIA-Linux-*.run’: No such file or directory
The file exists in the Downloads directory
David
This /path/to means your /path/to nvidia binary, like /home/[username]/Downloads/ or /tmp/ etc.
Hey, thanks for the article! I’ve been trying to get dual monitors working with gnome for the last 3 days and while this gets me closer (more options in nvidia-settings) BUT i still can’t get both screens working. I have two GT 610 GPU’s with the same monitor attached to each and they work fine with Noueveau. I’ve tried in Ubuntu 14.04/15.10 to no avail also. I now believe it is definitely the xorg.conf but i’ve tried every option i can think of? Using 1 x server, 2 x servers, multi-gpu off and on. Do you have any ideas? Looking at the nvidia-settings man page i found this “nvidia-xconfig -enable-all-gpus” but it doesn’t create a working xorg.conf either. Please help if you can!
Hi,
I have followed all the steps, but after final reboot I get the “Oh no…” screen.
the Xorg.0.log show message:
(EE) Failed to initialize GLX extension (Compatible NVIDIA X driver not found)
I was trying to update nvidia installer (with nvidia-installer –update), but with no success after reboot.
Is there any way to revert the changes made by installation?
Any help is appriciated.
Thank you!
[code]
$ nvidia-settings -v | grep version
nvidia-settings: version 358.16 ([email protected]) Mon Nov 16 19:58:01 PST 2015
[/code]
[code]
$ uname -a
Linux PC 4.2.6-301.fc23.x86_64 #1 SMP Fri Nov 20 22:22:41 UTC 2015 x86_64 x86_64 x86_64 GNU/Linux
[/code]
[code]
$ lspci | grep -i VGA
01:00.0 VGA compatible controller: NVIDIA Corporation GM204 [GeForce GTX 970] (rev a1)
02:00.0 VGA compatible controller: NVIDIA Corporation GM204 [GeForce GTX 970] (rev a1)
[/code]
Hi and thank you for the guide!
Unfortunately, I’ve run into a problem with the initramfs image on Fedora 23 and thought I’d share in case you have any suggestions. Essentially, at step 2.7 the initramfs image I generate while trying to blacklist the nouveau modules makes my system unbootable. I’ve tried a couple of things beyond the instructions as well, such as using the dracut option –omit-modules nouveau. The system always ends up hanging sometime after the “Journal” service starts up, though I doubt it’s related.
So the bind that I’m in is that the initramfs image I generate without nouveau drivers won’t boot.. and I can’t run the NVIDIA installer with an initramfs image that I can’t rmmod the nouveau modules while in use.
for reference, here’s me looking for recovery options:
https://www.reddit.com/r/Fedora/comments/3w6ehb/help_unable_to_boot_get_shell_after_attempted/
and here’s another fedora user referencing my thread with the same problem a few days later:
https://www.reddit.com/r/Fedora/comments/3wckgf/problem_with_nvidia_drivers/
Thanks again for the guide and any suggestions are appreciated.
Hi there. Following up on my own comment for anyone’s future reference. I was able to get past the boot hang-up by adding “edd=off” to my kernel boot parameters. This allowed me to boot into a nouveau-less initramfs and run the NVIDIA installer. I now have an X session with NVIDIA drivers. I’m still not sure what the issue is, but this resolved it.
My only residual problem is the NVIDIA installer failed with DKMS registration. Looking at the logs, it seems there’s a compiler version mismatch between my kernel (compiled with gcc 5.1) and current (gcc 5.3). Defining CC=gcc-5.1 didn’t help. So, I bailed on DKMS for the moment and just installed without it.
Yes i had the same issue as well and i had to install it without DKMS registration due to the mismatch gcc version that it was compiled. Will this affect anything else?
Hi Anastasios,
No effect to anything else. You can also upgrade your kernel to 4.18.15/4.18.16 to get it working without extra parameters.
[[email protected] ~]$ nvidia-installer -v|grep version
nvidia-installer: version 358.16 ([email protected])
[[email protected] ~]$ uname -a
Linux localhost.localdomain 4.2.6-301.fc23.x86_64 #1 SMP Fri Nov 20 22:22:41 UTC 2015 x86_64 x86_64 x86_64 GNU/Linux
[[email protected] ~]$ lspci |grep -i VGA
01:00.0 VGA compatible controller: NVIDIA Corporation GF106M [GeForce GTX 460M] (rev a1)
Hello,
I got to step 2.8.6 and the installation failed.
the output says something like this:
“failed to run ‘/usr/sbin/dmks build -m nvidia -v 352.63 -k 4.2.6-301.fc23.x86_64’:
Kernel preparation unnecessary for this kernel. Skipping…
Building Module:
Cleaning build area…
make Kernelrelese=4.2.6— Module
Kernel_Uname=4.2.6— ; make -C uvm module
Kernel_uname-4.2.6—
Kbuild_extmod=/var/lib/dkms/nvidia/352.63/build/uvm…(bad exit status:2)
Error! Bad return status for module build on kernel: 4.2.6—
Consult /var/lib/dkms/nvidia/352.63/build/make.log for more info”
Any ideas?
Thanks.
Hi Nick,
Could you post output of following commands:
after step 2.8.6 I get:
Error: Unable to load the ‘nvidia’ kernel module.
Error: Installation has failed. Please see the file ‘/var/lo/nvidia-installer.log’ for details
secure boot was enabled on my Asus Z97-A motherboard…
It is all fixed now.
nvidia-installer: version 340.96 ([email protected])
Linux localhost.localdomain 4.2.7-200.fc22.x86_64 #1 SMP Thu Dec 10 03:28:47 UTC 2015 x86_64 x86_64 x86_64 GNU/Linux
01:00.0 VGA compatible controller: NVIDIA Corporation GT200 [GeForce GTX 260] (rev a1)
** updating dnf gave me gcc-5.3.*; the kernel was compiled with gcc-5.1*. The nvidia installer gave error and asked if I wished to abort. I downgraded the cc suite and all worked fine.. The only exception was that the installer never gave me the DKMS option. Thank you for posting this guide!
wow thanks for this guide,
really helped me out as i just did a Fedora 23 Workstation with Nvidia GTX 970 as my GPU
#nvidia-installer -v |grep version
nvidia-installer: version 358.16 ([email protected])
# uname -a
Linux ivy 4.2.7-300.fc23.x86_64 #1 SMP Wed Dec 9 22:28:30 UTC 2015 x86_64 x86_64 x86_64 GNU/Linux
# lspci |grep -i VGA
01:00.0 VGA compatible controller: NVIDIA Corporation GM204 [GeForce GTX 970] (rev a1)
All i wanna say is that I updated fedora to kernel 4.2.7 and downloaded a 358.16 version of drivers and it didn’t work (i couldnt even load my system after that), I tried the same thing with 352.63 version and everything went smooth after that with one exception –
“Unable to perform the runtime configuration check for 32-bit library ‘libEGL.so.1’ ….; this is typically caused by the lack of a 32-bit compatibility environment. Assume successful installation”
JR,
I was following your steps until 2.8.6 where DKMS failed in the driver installation. It appears that the driver was installed so I rebooted but was unable to login. A dialog box came up before any login screen stating that it was unable to recover.
make log: http://pastebin.com/x5ci8nxs
nvidia installer log: http://pastebin.com/JM7c3y7P
kernel-4.2.3-300.fc23.x86_64
kernel-4.2.7-300.fc23.x86_64
kernel-core-4.2.3-300.fc23.x86_64
kernel-core-4.2.7-300.fc23.x86_64
kernel-devel-4.2.3-300.fc23.x86_64
kernel-devel-4.2.7-300.fc23.x86_64
kernel-modules-4.2.3-300.fc23.x86_64
kernel-modules-4.2.7-300.fc23.x86_64
Hi Kyle,
Thanks for reporting this. This should be fixed on next kernel version, when it’s build using gcc 5.3.
And yes, it looks like that kernel 4.2.7 is build using gcc 5.1.1:
What you can do is set IGNORE_CC_MISMATCH environment variable to “1” or downgrade gcc to 5.1.1 version.
Btw. You should also have kernel-headers-4.2.7-300.fc23.x86_64 installed, if you followed this guide.
Yep that worked for me, thanks!
Worked for me. Thanks!
nvidia-installer: version 358.16 ([email protected]) Mon Nov 16 19:59:22 PST 2015
Linux darwin 4.2.7-300.fc23.x86_64 #1 SMP Wed Dec 9 22:28:30 UTC 2015 x86_64 x86_64 x86_64 GNU/Linux
01:00.0 VGA compatible controller: NVIDIA Corporation GF108 [GeForce GT 430] (rev a1)
Thanks JR, IGNORE_CC_MISMATCH environment variable to “1” worked perfectly…