Black screen on boot | F40 | NVIDIA GPU

Ask Fedora
, , ,
1

Hello,

I’m facing some issues using nvidia driver after a clean install of Fedora 40. After I install the driver and I reboot, I am facing a black screen (no signal). The only way after that to get some image is to reboot and remove the options in grub blacklisting nouveau.

I have tried reinstalling the drivers multiple times, and decided to do a clean install before asking for help. I’ll try to detail my system and the steps taken in hopes someone wiser can help me figure out what is wrong :slight_smile:

Output of inxi -Fzxx

System:
  Kernel: 6.9.9-200.fc40.x86_64 arch: x86_64 bits: 64 compiler: gcc
    v: 2.41-37.fc40
  Desktop: GNOME v: 46.3.1 tk: GTK v: 3.24.43 wm: gnome-shell dm: GDM
    Distro: Fedora Linux 40 (Workstation Edition)
Machine:
  Type: Desktop System: ASUS product: All Series v: N/A
    serial: <superuser required>
  Mobo: ASUSTeK model: Z97-PRO GAMER v: Rev X.0x
    serial: <superuser required> part-nu: All UEFI: American Megatrends v: 2203
    date: 02/26/2016
CPU:
  Info: quad core model: Intel Core i7-4790K bits: 64 type: MT MCP
    arch: Haswell rev: 3 cache: L1: 256 KiB L2: 1024 KiB L3: 8 MiB
  Speed (MHz): avg: 1737 high: 4400 min/max: 800/4400 cores: 1: 800 2: 800
    3: 800 4: 1500 5: 800 6: 4001 7: 800 8: 4400 bogomips: 63966
  Flags: avx avx2 ht lm nx pae sse sse2 sse3 sse4_1 sse4_2 ssse3 vmx
Graphics:
  Device-1: NVIDIA GM204 [GeForce GTX 970] vendor: ASUSTeK driver: nouveau
    v: kernel arch: Maxwell pcie: speed: 2.5 GT/s lanes: 16 ports:
    active: DVI-I-1 empty: DP-1,DVI-D-1,HDMI-A-1 bus-ID: 01:00.0
    chip-ID: 10de:13c2 temp: 47.0 C
  Display: wayland server: X.org v: 1.20.14 with: Xwayland v: 24.1.1
    compositor: gnome-shell driver: gpu: nouveau display-ID: 0
  Monitor-1: DVI-I-1 model: Asus VS278 res: 1920x1080 dpi: 82
    diag: 686mm (27")
  API: OpenGL v: 4.3 vendor: mesa v: 24.1.4 glx-v: 1.4 es-v: 3.2
    direct-render: yes renderer: NV124 device-ID: 10de:13c2 display-ID: :0.0
  API: EGL Message: EGL data requires eglinfo. Check --recommends.
Audio:
  Device-1: Intel 9 Series Family HD Audio vendor: ASUSTeK
    driver: snd_hda_intel v: kernel bus-ID: 00:1b.0 chip-ID: 8086:8ca0
  Device-2: NVIDIA GM204 High Definition Audio vendor: ASUSTeK
    driver: snd_hda_intel v: kernel pcie: speed: 2.5 GT/s lanes: 16
    bus-ID: 01:00.1 chip-ID: 10de:0fbb
  API: ALSA v: k6.9.9-200.fc40.x86_64 status: kernel-api
  Server-1: JACK v: 1.9.22 status: off
  Server-2: PipeWire v: 1.0.7 status: active with: 1: pipewire-pulse
    status: active 2: wireplumber status: active 3: pipewire-alsa type: plugin
Network:
  Device-1: Intel Ethernet I218-V vendor: ASUSTeK driver: e1000e v: kernel
    port: f040 bus-ID: 00:19.0 chip-ID: 8086:15a1
  IF: eno1 state: up speed: 1000 Mbps duplex: full mac: <filter>
Drives:
  Local Storage: total: 2.04 TiB used: 11.17 GiB (0.5%)
  ID-1: /dev/sda vendor: Western Digital model: WD1003FZEX-00K3CA0
    size: 931.51 GiB speed: 6.0 Gb/s serial: <filter>
  ID-2: /dev/sdb vendor: Kingston model: SHFS37A240G size: 223.57 GiB
    speed: 6.0 Gb/s serial: <filter>
  ID-3: /dev/sdc vendor: Maxtor model: STM31000528AS size: 931.51 GiB
    speed: 3.0 Gb/s serial: <filter>
Partition:
  ID-1: / size: 221.98 GiB used: 10.79 GiB (4.9%) fs: btrfs dev: /dev/sdb3
  ID-2: /boot size: 973.4 MiB used: 362.2 MiB (37.2%) fs: ext4
    dev: /dev/sdb2
  ID-3: /boot/efi size: 598.8 MiB used: 19 MiB (3.2%) fs: vfat
    dev: /dev/sdb1
  ID-4: /home size: 221.98 GiB used: 10.79 GiB (4.9%) fs: btrfs
    dev: /dev/sdb3
Swap:
  ID-1: swap-1 type: zram size: 8 GiB used: 0 KiB (0.0%) priority: 100
    dev: /dev/zram0
Sensors:
  System Temperatures: cpu: 34.0 C mobo: N/A gpu: nouveau temp: 47.0 C
  Fan Speeds (rpm): N/A gpu: nouveau fan: 0
Info:
  Memory: total: 16 GiB available: 15.55 GiB used: 2.53 GiB (16.3%)
  Processes: 849 Power: uptime: 7m wakeups: 0 Init: systemd v: 255
    target: graphical (5) default: graphical
  Packages: pm: flatpak pkgs: 2 Compilers: N/A Shell: Bash v: 5.2.26
    running-in: gnome-terminal inxi: 3.3.34

Installed drivers following this guide Howto/NVIDIA - RPM Fusion

First checked for secure boot
mokutil --sb-state returned SecureBoot disabled

Then I installed the drivers

sudo dnf update -y 
sudo dnf install akmod-nvidia
sudo dnf install xorg-x11-drv-nvidia-cuda

and then waited until modinfo returned the nvidia driver version before rebooting

modinfo -F version nvidia
555.58.02

After reboot then screen is black (no signal)

Here is the journal for the boot when the monitor gets no signal
https://paste.centos.org/view/7987d76e

As well as the journal from when driver was installed
https://paste.centos.org/view/b6fe8f6d

Any idea what might be causing this? Any more information I can provide?

thank you for your time,
Regards

Andre

1 Like
2

Removed audio, flatpak, gnome, pipewire, server, wayland, wireplumber

3

Looking at those 2 logs it would appear that the second was the initial installation of the drivers, and that you may have booted early before the driver was properly built.

The first seems to show the driver being loaded, though there may have been some part of the build of the driver still in progress during the boot.

Please run sudo akmods --force --rebuild
Wait until that completes then after waiting another minute or so reboot again and let us know if there is any difference.

4

Added workstation and removed intel

5

Added wayland

6

I ran the command and it finished successfully. Then I waited 5 minutes before rebooting.
Unfortunately the result was the same.

Here is the journal of the boot following the rebuild
https://paste.centos.org/view/45203b3f

7

What happens if you add nomodeset into the grub kernel command line?
or nvidia_drm.modeset=0?

8

I tried both options in different boots.

nomodeset
https://paste.centos.org/view/40629aa2

nvidia_drm.modeset=0
https://paste.centos.org/view/bf5464ab

This time in both tries I could see for a brief moment the fedora logo and loading animation in a lower resolution, but after that the screen didn’t have any signal.

Also in both journals I the nvidia errors regarding nvidia-powerd seem to be gone.

Should I keep one of these options permanently?

9

Please provide the full archive generated by nvidia-bug-report:
https://rpmfusion.org/Howto/NVIDIA#Bug_Report
(eventually from a remote computer).

Please don’t mess with dracut rebuild, we have an option to avoid bundling the nvidia driver into dracut, so this dracut command is a NOP.

1 Like
10

Here is the nvidia-bug-report:
https://paste.centos.org/view/bd5b1315

Since I couldn’t get any image when I’m using the nvidia driver I used crontab to run the script after reboot. It seemed to have worked, but let me know if it has all the information.

11

Something is forcing nouveau to be loaded despite that nouveau is supposed to be blocked from boot.

Is there any files in /etc/modprobe.d or /etc/modules-load.d (and related directories in /usr/lib ?)

As the nvidia driver from rpmfusion is concerned, the blocklist in /proc/cmdline is what is needed and sufficient to block nouveau from loading. But if some script is enforcing nouveau to be loaded before

Also it seems weird that nvidia-fallback didn’t detected the nouveau blocklist from /proc/cmdline is this standard Fedora ?

12

/etc/modprobe.d

appletalk-blacklist.conf     rds-blacklist.conf
atm-blacklist.conf           rose-blacklist.conf
ax25-blacklist.conf          sch_choke-blacklist.conf
batman-adv-blacklist.conf    sch_drr-blacklist.conf
firewalld-sysctls.conf       sch_etf-blacklist.conf
floppy-blacklist.conf        sch_gred-blacklist.conf
kvm.conf                     sch_mqprio-blacklist.conf
l2tp_eth-blacklist.conf      sch_multiq-blacklist.conf
l2tp_ip-blacklist.conf       sch_netem-blacklist.conf
l2tp_netlink-blacklist.conf  sch_qfq-blacklist.conf
l2tp_ppp-blacklist.conf      sch_red-blacklist.conf
lockd.conf                   sch_sfb-blacklist.conf
netrom-blacklist.conf        sctp-blacklist.conf
nfc-blacklist.conf           vhost.conf
nvdimm-security.conf

/etc/modules-load.d is empty.

/usr/lib/modules-load.d

fuse-overlayfs.conf  fwupd-msr.conf  open-vm-tools.conf  podman-iptables.conf

/usr/lib/modprobe.d

dist-alsa.conf       nvidia-power-management.conf  openfwwf.conf  systemd.conf
dist-blacklist.conf  nvidia-uvm.conf               README

This is Fedora 40 workstation. I did a clean install to make sure there wasn’t anything else causing issues before asking for help. Not sure if that was the question or if you were asking if this is normal in Fedora :smiley:

13

The nvidia-but-report has:

Jul 21 15:41:22 fedora /usr/bin/nvidia-powerd[888]: Found unsupported configuration. Exiting...

and

/sys/bus/pci/devices/0000:01:00.0/power/runtime_usage does not exist

Many vendors have “BIOS” firmware updates that include updates to power management. You should use fwupdtool or check the vendor’s site for firmware updates.

Your model has multiple probes in the LHDB. Scanning the probes may provide useful details in user comments, and you may find that some users have success with newer video cards.

14

Thank you for your reply.

I checked with fwupdtool and in the vendor’s site and I already had the latest motherboard BIOS version (from 31/03/2016).

I’ll take a look into this see if I can find some useful information.

Another point I forgot to mention and might be relevant, I was previously running Fedora 39 with an older nvidia driver (maybe 550, not sure) and I never had any of these problems. I decided to do a clean install when upgrading to 40 to avoid any issues :smiley:

15

That is perfectly normal. It is merely a message telling the user that the nvidia GPU in use is not supported by nvidia-powerd (there are many).

1 Like
16

Might be related to Installing NVIDIA drivers causes Black Screen on Bootup | F40 | - #3 by manuelmorales ?

1 Like