Nvidia installation chaos (F44 KDE Plasma)

Hi! I’m new to Linux. I try to be moderately tech savvy, but it’s been hard to gain a solid understanding so far due to being literally unable to use my system. Apologies if I’m a little slow to understand any help provided, but thanks in advance!

I can’t get Nvidia drivers to work. I’ve been troubleshooting this issue for several days. I’ve come across many similar forum posts, but none of the advice provided seems to be of use for my circumstances.

My GPU is an RTX 3060 ti, so I have been installing the most up to date versions of everything as per rpm fusion’s recommendations.

Things I’ve tried -

  • Reinstall Fedora on top of itself
  • Wipe partition and reinstall Fedora completely
  • Download and flash Fedora again in case of corrupted files
  • Manually making sure every necessary rpm repo is available and being used
  • Making sure secure boot is 1000% disabled (confirmed in both BIOS and Konsole)
  • Waiting an entire hour after completing the driver install, to be sure five minutes has passed
  • Using the Discover GUI to install instead of Konsole

Things I’ve experienced -

  • Booting to black screen, able to access console with CTRL+ALT+F3
  • Booting to black screen, completely unresponsive
  • Booting to console (unprompted), sometimes able to input commands, sometimes unable to type
  • Booting, but only one of my monitors functions and is low resolution
  • Booting seemingly normally, but after login black screen with only cursor available (my most recent fail state)

I’m at a total loss at this point. I’ve been doing my best to fix it myself and have learned a fair bit in doing so, but I’m now out of ideas and don’t know what to try next, please help me :weary_face:

In the current state of your system, are you still able to get to a console with CTRL+ALT+F3?

Hi, yes

OK, good, so at least it’s possible to look at some logs and try to figure out what’s going on.

As a quick first step, what do you see from these commands in the terminal?

sudo dmesg | grep -i nvidia

sudo journalctl -b 0 -g nvidia

There’s more than I could quickly copy by hand, is there a way I can get the console output available to post here in full?

There are a few returns from the journalctl command that return with fail messages

Copy and paste is the easiest method as pre-formatted text (</> button in this dialog box when you’re writing as post… you can surround it with three back ticks on blank line, top and bottom, too

like this

If you don’t use pre-formatted text, it buggers up the formatting when anyone quotes it.

However if there’s too much to do that comfortably run those commands piped into fpaste and paste the URL(s) it provides here.

eg.

sudo dmesg | grep -i nvidia | fpaste
sudo journalctl -b 0 -g nvidia | fpaste

Of course you can always direct the output to a file and then use an editor to trim it up before copy and pasting the entire thing into here. There’s a post limit of 32 KB.

Thank you!

These are the results:

dmesg | grep -i nvidia

journatlctl -b 0 -g nvidia

[    6.451169] input: HDA NVidia HDMI/DP,pcm=3 as /devices/pci0000:00/0000:00:1c.0/0000:02:00.1/sound/card1/input12
[    6.451493] input: HDA NVidia HDMI/DP,pcm=7 as /devices/pci0000:00/0000:00:1c.0/0000:02:00.1/sound/card1/input13
[    6.451536] input: HDA NVidia HDMI/DP,pcm=8 as /devices/pci0000:00/0000:00:1c.0/0000:02:00.1/sound/card1/input14
[    6.451553] input: HDA NVidia HDMI/DP,pcm=9 as /devices/pci0000:00/0000:00:1c.0/0000:02:00.1/sound/card1/input15
[   33.318189] nouveau 0000:02:00.0: NVIDIA GA104 (b74000a1)
Jul 03 10:25:11 rinux kernel: input: HDA NVidia HDMI/DP,pcm=3 as /devices/pci0000:00/0000:00:1c.0/0000:02:00.1/sound/card1/input12
Jul 03 10:25:11 rinux kernel: input: HDA NVidia HDMI/DP,pcm=7 as /devices/pci0000:00/0000:00:1c.0/0000:02:00.1/sound/card1/input13
Jul 03 10:25:11 rinux kernel: input: HDA NVidia HDMI/DP,pcm=8 as /devices/pci0000:00/0000:00:1c.0/0000:02:00.1/sound/card1/input14
Jul 03 10:25:11 rinux kernel: input: HDA NVidia HDMI/DP,pcm=9 as /devices/pci0000:00/0000:00:1c.0/0000:02:00.1/sound/card1/input15
Jul 03 10:25:12 rinux systemd[1]: Started nvidia-powerd.service - nvidia-powerd service.
Jul 03 10:25:12 rinux audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=nvidia-powerd comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Jul 03 10:25:12 rinux nvidia-powerd[1079]: nvidia-powerd version:2.0 (build 1)
Jul 03 10:25:12 rinux systemd[1]: nvidia-powerd.service: Deactivated successfully.
Jul 03 10:25:12 rinux audit[1]: SERVICE_STOP pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=nvidia-powerd comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Jul 03 10:25:30 rinux akmods[1061]: Building and installing nvidia-kmod [FAILED]
Jul 03 10:25:35 rinux akmods[1061]: Building rpms failed; see /var/cache/akmods/nvidia/595.80-1-for-7.0.14-201.fc44.x86_64.failed.log for details
Jul 03 10:25:37 rinux systemd[1]: Starting nvidia-fallback.service - Fallback to nouveau as nvidia did not load...
Jul 03 10:25:38 rinux kernel: nouveau 0000:02:00.0: NVIDIA GA104 (b74000a1)
Jul 03 10:25:40 rinux systemd[1]: Finished nvidia-fallback.service - Fallback to nouveau as nvidia did not load.
Jul 03 10:25:40 rinux audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=nvidia-fallback comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Jul 03 10:28:08 rinux systemd[12565]: Starting app-nvidia\x2dsettings\x2duser@autostart.service - nvidia-settings...
Jul 03 10:28:08 rinux systemd[12565]: Started app-nvidia\x2dsettings\x2duser@autostart.service - nvidia-settings.
Jul 03 10:28:08 rinux nvidia-settings[13095]: ERROR: NVIDIA driver is not loaded
Jul 03 10:28:08 rinux systemd[12565]: app-nvidia\x2dsettings\x2duser@autostart.service: Main process exited, code=exited, status=1/FAILURE
Jul 03 10:28:08 rinux systemd[12565]: app-nvidia\x2dsettings\x2duser@autostart.service: Failed with result 'exit-code'.
Jul 03 10:28:08 rinux systemd[12565]: app-nvidia\x2dsettings\x2duser@autostart.service: Consumed 129ms CPU time over 161ms wall clock time, 65.9M memory peak.
Jul 03 11:00:55 rinux sudo[15698]:      rin : TTY=tty3 ; PWD=/home/rin ; USER=root ; COMMAND=/usr/sbin/journalctl -b 0 -g nvidia
Jul 03 11:11:17 rinux sudo[16295]:      rin : TTY=tty3 ; PWD=/home/rin ; USER=root ; COMMAND=/usr/sbin/journalctl -b 0 -g nvidia
Jul 03 10:25:35 rinux akmods[1061]: Building rpms failed; see /var/cache/akmods/nvidia/595.80-1-for-7.0.14-201.fc44.x86_64.failed.log for details
Jul 03 10:25:37 rinux systemd[1]: Starting nvidia-fallback.service - Fallback to nouveau as nvidia did not load...
Jul 03 10:25:38 rinux kernel: nouveau 0000:02:00.0: NVIDIA GA104 (b74000a1)
Jul 03 10:25:40 rinux systemd[1]: Finished nvidia-fallback.service - Fallback to nouveau as nvidia did not load.
Jul 03 10:25:40 rinux audit[1]: SERVICE_START pid=1 uid=0 auid=4294967295 ses=4294967295 subj=system_u:system_r:init_t:s0 msg='unit=nvidia-fallback comm="systemd" exe="/usr/lib/systemd/systemd" hostname=? addr=? terminal=? res=success'
Jul 03 10:28:08 rinux systemd[12565]: Starting app-nvidia\x2dsettings\x2duser@autostart.service - nvidia-settings...
Jul 03 10:28:08 rinux systemd[12565]: Started app-nvidia\x2dsettings\x2duser@autostart.service - nvidia-settings.
Jul 03 10:28:08 rinux nvidia-settings[13095]: ERROR: NVIDIA driver is not loaded

OK, pretty clear then that the driver install failed. (Well, the RPMFusion packages were installed but they failed to build the kernel module.)

Can you upload its build log?

sudo cat /var/cache/akmods/nvidia/595.80-1-for-7.0.14-201.fc44.x86_64.failed.log | fpaste

Results

See man fpaste for a quick way to transfer output from terminal commands and the post the URL here. For maximum visibility you can then use a system with working GUI to post the output (as pre-formatted web-discoverable text using the </> button) here. If you don’t have easy access to a system with a web browser you can use a Live USB installer to view the journals (using journalctl with the —directory option) and paste directly into a forum post.

This is full of internal compiler errors.

In file included from /usr/src/kernels/7.0.14-201.fc44.x86_64/include/linux/hrtimer.h:21,
                 from /usr/src/kernels/7.0.14-201.fc44.x86_64/include/linux/interrupt.h:13,
                 from ././common/inc/nv-time.h:30,
                 from nvidia-uvm/uvm_linux.h:37,
                 from nvidia-uvm/uvm_common.h:43,
                 from nvidia-uvm/uvm_api.h:28,
                 from nvidia-uvm/uvm_va_space.c:24:
/usr/src/kernels/7.0.14-201.fc44.x86_64/include/linux/timer.h:192:50: internal compiler error: in c_addr_space_name, at c-family/c-common.cc:657

Indeed, looks pretty unpleasant :slightly_frowning_face:

The other puzzle (in common with various other recent bug reports) is why falling back to nouveau doesn’t work properly. KDE Plasma clearly does in principle work with nouveau, because that’s how all our live USBs work.

Ah good idea, I’ll do this for future replies. Thanks those who were transcribing my links!

I’ll have to admit, I blacklisted nouveau about 3 years ago and haven’t used it at all to my knowledge, so when it buggers up I’m lost.

I’d start from first principles here, given that the OP has had to fiddle quite a bit so perhaps removing akmods and reinstalling the lot from the RPM Fusion guide might be apposite.

Clean slate and all that,

I’ve not heard of any gcc 16 internal errors from anyone else building this stuff, but then again I’ve not been particularly watching gcc bug reports either!

Certainly worth a try.

@rinux , is the CPU overclocked?

@rinux the results of the attempt by akmods to build and create the drives and insert them into your system will be documented in /var/cache/akmods/nvidia/.last.log

It should start like this:

2026/06/05 21:58:50 akmods: Building RPM using the command '/usr/sbin/akmodsbuild --kernels 7.0.10-201.fc44.x86_64 /usr/src/akmods/nvidia-kmod.latest'
/tmp/akmods.jPnPQzjk/results/kmod-nvidia-7.0.10-201.fc44.x86_64-595.80-1.fc44.x86_64.rpm
2026/06/05 21:59:19 akmodsbuild: * Rebuilding /usr/src/akmods/nvidia-kmod.latest for kernel(s) 7.0.10-201.fc44.x86_64: Building target platforms: x86_64
2026/06/05 21:59:19 akmodsbuild: Building for target x86_64
2026/06/05 21:59:19 akmodsbuild: Installing /usr/src/akmods/nvidia-kmod.latest
2026/06/05 21:59:19 akmodsbuild: setting SOURCE_DATE_EPOCH=1780012800
2026/06/05 21:59:19 akmodsbuild: Executing(%mkbuilddir): /bin/sh -e /var/tmp/rpm-tmp.PXKMKN
2026/06/05 21:59:19 akmodsbuild: Executing(%prep): /bin/sh -e /var/tmp/rpm-tmp.MEAaim
2026/06/05 21:59:19 akmodsbuild: + umask 022

and it absolutely must end looking like this:

2026/06/05 21:59:19 akmods: Installing newly built rpms
2026/06/05 21:59:19 akmods: DNF detected
Updating and loading repositories:
Repositories loaded.
Package                                         Arch   Version            Repository        Size
Upgrading:
 kmod-nvidia-7.0.10-201.fc44.x86_64             x86_64 3:595.80-1.fc44    @commandline  10.5 MiB
   replacing kmod-nvidia-7.0.10-201.fc44.x86_64 x86_64 3:595.71.05-1.fc44 @commandline  10.5 MiB

Transaction Summary:
 Upgrading:          1 package
 Replacing:          1 package

Total size of inbound packages is 9 MiB. Need to download 0 B.
After this operation, 7 KiB will be freed (install 10 MiB, remove 10 MiB).
Running transaction
[1/4] Verify package files              100% |  38.0   B/s |   1.0   B |  00m00s
[2/4] Prepare transaction               100% |   4.0   B/s |   2.0   B |  00m00s
[3/4] Upgrading kmod-nvidia-7.0.10-201. 100% |   1.3 MiB/s |  10.5 MiB |  00m08s
[4/4] Removing kmod-nvidia-7.0.10-201.f 100% |   0.0   B/s |   7.0   B |  00m08s
Warning: skipped OpenPGP checks for 1 package from repository: @commandline
Complete!
2026/06/05 21:59:36 akmods: Successful.

If you don’t get output which finishes with “Successful” and a few lines about rpm’s being installed, it won’t be working when you reboot.

Quite why nouveau is also unhappy is something I’ll leave to those who have used it. I can barely spell it…

Ooh it might be? I haven’t done any overclocking in a long time but I do remember messing with stuff when I first bought the thing years ago. I’ll check the BIOS in a sec

I’m sorry, how do I view this log?