There actually are several other posts concerning this issue, but everyone’s case is a bit unique so I decided to start a new thread rather than piggyback off someone else’s thread. I have a desktop which was running Fedora 40, it had upgraded from 38 to 39 and then to 40 and it was running great. Then there was some update within 40 and now it boots up to my login screen, I click my username & log in. Screen flashes like it’s loading, but nothing ever comes up. Just a black screen. I can see and move the mouse curser around though. I can also switch to F2 and log in via the terminal with no problem, but startx just brings me back to the same black screen with the mouse curser.
The nomodeset option does nothing. I already tried something else too but don’t remember what it was - I multiboot and was busy with other stuff so I just left Fedora for a while when I couldn’t get it to work. The log files in .local/share/xorg are useless, the last one there was from the last successful GUI login at the beginning of June. I did retrieve a boot log from the var directory. Current kernel is the 6.9.8-200, but the kernel itself has been updated since the issue started. I remember reading something about the nvidia stuff and I did try some things but they obviously didn’t work, and as I said before I don’t remember exactly what stuff I tried so more than happy to try something again. Not sure how to upload the boot log file. I could just paste the contents, but it’s pretty long and … I didn’t see anything there except a lot of lines with 39 … I’m on 40. But I am not sure that’s what the 39 actually meant as they all seemed to have that in. And the computer booted into 40 just fine for a while.
I am currently running another linux distro on the same computer, so I can access any file you need me to. Running things directly from Fedora I can snap a pic with my cellphone & upload it back here.
Okay, lastly, my CPU is a W-2145, I have 64GB RAM, 2 regular HDs inside but Fedora is booting off a third drive, an M2 SSD. Video card is a RTX-3060-Lite, there are no integrated graphics. My system boots legacy MBR not the new EFI. If there is any other relevant information I should have included please let me know. Thank you in advance!
Sadly still no replies. So just not really using Fedora anymore, sadly. I guess maybe due to the much larger user base Debian flavors have more support and work the bugs out better or faster. Really trying to get going with another Linux OS non-Debian based, had high hopes for Fedora due to it’s working with DaVinci Resolve, but even that won’t work without a graphical GUI. I will try again in a few more weeks probably. If I do find a solution I will log back in here and post it. Thank you & have a great week!
Hi there. If you’re still willing to give it a shot, first thing you could try is test if the issue still persists with the latest kernel 6.9.11
. There are indeed several posts with nvidia and blank screen issues, some of them got solved with the latest kernel update. You should be able to do that from a tty console.
If the issue is still there, please provide the following:
- Is it working with nvidia and X11?
- Is it working with nouveau and Wayland?
- After reboot, logging in with issues and opening console, please provide the journal output for the last boot. That can be done with
journalctl -b | fpaste --raw-url
, and posting here the generated link. - From the same session, run
inxi -Fzxx
. Post the contents here as preformatted text</>
. You can pipe the output tofpaste
, as above (given that you’re in a console), but this time post the output here, as it is not that long. - Post the output of
sudo dnf list installed \*nvidia\*
.
Hello! First thank you very much for helping!!! Now I am the one who must apologize for taking so long. These links are hopefully still good… So I have two links here:
https://paste.centos.org/view/raw/0f980787
https://paste.centos.org/view/raw/07d2e265
Those are the outputs of inxi and journalctl, I sorta forgot about the nvidia but will go back and run that too. Not sure about the other questions though… It WAS working fine, now it’s not, period. As far as nouveau, I confess I have not tried to go back and switch it again, though my main purpose specifically here would be Resolve which will only run on Nvidia hence the reason I updated the card in the first place. Is there an easy way to switch back to nouveau, even just temporary? Maybe from the boot line in Grub … I did find a boot line in one of the forum questions, I was trying to force nvidia though not nouveau, but it kept giving me errors when remaking grub so I abandoned it. I will also repost here with my current grub entry, which I should have here. But in the meantime here are the the two log files with lots of info in one of them, Thank you again, very very much. I will post again soon with the additional info.
- Peter
Those links are dead. Centos only keeps the pasted data for 24 hours so you would need to rerun the command to upload the data then immediately paste it here.
Note that for inxi it is much preferable to paste the text into your post (using the preformatted text tags) since the length is reasonable (as will be the output of the nvidia command when you post it)
If you peruse the journalctl output and select only the relevant parts then it can be pasted here the same way.
Okay, just got the links so they should be good for a uh, day. Anyway, here is the journal link:
https://paste.centos.org/view/raw/936e9e83
I will attempt to post the other info here, though it’s possible I could leave out some info that could be relevant so here are the fpaste links just in case for the raw data:
https://paste.centos.org/view/raw/cd6173c4
https://paste.centos.org/view/raw/11064cbb
Okay, here is the info here:
System:
Kernel: 6.9.11-200.fc40.x86_64 arch: x86_64 bits: 64 compiler: gcc v: 2.41-37.fc40
Graphics:
Device-1: NVIDIA GA106 [GeForce RTX 3060 Lite Hash Rate] vendor: Gigabyte driver: nvidia
v: 555.58.02 arch: Ampere pcie: speed: 2.5 GT/s lanes: 16 ports: active: none off: DP-1, DP-2,
HDMI-A-1, HDMI-A-2 empty: none bus-ID: 65:00.0 chip-ID: 10de:2504
Device-2: Microdia Integrated Camera driver: uvcvideo type: USB rev: 2.0 speed: 480 Mb/s
lanes: 1 bus-ID: 1-2:2 chip-ID: 0c45:64ab
Device-3: Realtek Web Camera driver: snd-usb-audio,uvcvideo type: USB rev: 2.1 speed: 480 Mb/s
lanes: 1 bus-ID: 4-2.4:7 chip-ID: 0bda:2137
Device-4: Logitech CrystalCam driver: snd-usb-audio,uvcvideo type: USB rev: 2.0
speed: 480 Mb/s lanes: 1 bus-ID: 4-7:3 chip-ID: 046d:0894
Display: server: X.org v: 1.20.14 with: Xwayland v: 24.1.1 compositor: gnome-shell driver:
gpu: nvidia,nvidia-nvswitch note: X driver n/a, try sudo/root tty: 160x45
Monitor-1: DP-1 model: HP V24i G5 res: 1920x1080 dpi: 93 diag: 604mm (23.8")
Monitor-2: DP-2 model: Envision G917w1 res: 1440x900 dpi: 89 diag: 484mm (19.1")
Monitor-3: HDMI-A-1 model: MACROSILICON res: 1280x720 dpi: 61 diag: 604mm (23.8")
Monitor-4: HDMI-A-2 model: Samsung SyncMaster res: 1280x1024 dpi: 90 diag: 450mm (17.7")
API: OpenGL Message: GL data unavailable in console. Try -G --display
API: EGL Message: EGL data unavailable in console, eglinfo missing.
Info:
Memory: total: 64 GiB available: 62.38 GiB used: 1.83 GiB (2.9%)
Processes: 335 Power: uptime: 13m wakeups: 0 Init: systemd v: 255 target: graphical (5)
default: graphical
Packages: Compilers: clang: 17 gcc: 14.1.1 Shell: Bash v: 5.2.26 running-in: tty 2
inxi: 3.3.34
Installed Packages
akmod-nvidia.x86_64 3:555.58.02-1.fc40 @rpmfusion-nonfree-nvidia-driver
kmod-nvidia-6.9.11-200.fc40.x86_64.x86_64 3:555.58.02-1.fc40 @@commandline
kmod-nvidia-6.9.4-200.fc40.x86_64.x86_64 3:555.58.02-1.fc40 @@commandline
kmod-nvidia-6.9.8-200.fc40.x86_64.x86_64 3:555.58.02-1.fc40 @@commandline
libva-nvidia-driver.x86_64 0.0.12-2.fc40 @updates
nvidia-gpu-firmware.noarch 20240709-1.fc40 @updates
nvidia-modprobe.x86_64 3:555.58.02-1.fc40 @rpmfusion-nonfree-nvidia-driver
nvidia-persistenced.x86_64 3:555.58.02-1.fc40 @rpmfusion-nonfree-nvidia-driver
nvidia-query-resource-opengl.x86_64 1.0.0-17.fc40 @fedora
nvidia-query-resource-opengl-lib.x86_64 1.0.0-17.fc40 @fedora
nvidia-settings.x86_64 3:555.58.02-1.fc40 @rpmfusion-nonfree-nvidia-driver
nvidia-texture-tools.x86_64 2.1.2-9.fc40 @fedora
nvidia-xconfig.x86_64 3:555.58.02-1.fc40 @rpmfusion-nonfree-nvidia-driver
xorg-x11-drv-nvidia.x86_64 3:555.58.02-1.fc40 @rpmfusion-nonfree-nvidia-driver
xorg-x11-drv-nvidia-cuda.x86_64 3:555.58.02-1.fc40 @rpmfusion-nonfree-nvidia-driver
xorg-x11-drv-nvidia-cuda-libs.i686 3:555.58.02-1.fc40 @rpmfusion-nonfree-nvidia-driver
xorg-x11-drv-nvidia-cuda-libs.x86_64 3:555.58.02-1.fc40 @rpmfusion-nonfree-nvidia-driver
xorg-x11-drv-nvidia-devel.x86_64 3:555.58.02-1.fc40 @rpmfusion-nonfree-nvidia-driver
xorg-x11-drv-nvidia-kmodsrc.x86_64 3:555.58.02-1.fc40 @rpmfusion-nonfree-nvidia-driver
xorg-x11-drv-nvidia-libs.i686 3:555.58.02-1.fc40 @rpmfusion-nonfree-nvidia-driver
xorg-x11-drv-nvidia-libs.x86_64 3:555.58.02-1.fc40 @rpmfusion-nonfree-nvidia-driver
xorg-x11-drv-nvidia-power.x86_64 3:555.58.02-1.fc40 @rpmfusion-nonfree-nvidia-driver
Hope that’s all readable. Thank you so much for your help and patience! Everything else is the same as previous reply. Oh, and I STILL forgot to copy the GRUB entry! I WILL get that! Maybe you won’t need it though but all the same. Thank you again!
Okay back again with my current GRUB boot line:
menuentry "Fedora Linux 40"{
insmod part_msdos
insmod btrfs
insmod ext2
search --no-floppy --fs-uuid --set=root e09f7118-34ed-49ba-82f4-74b481b8a674
linux /boot/vmlinuz-6.9.11-200.fc40.x86_64 root=/dev/nvme0n1p1
initrd /boot/initramfs-6.9.11-200.fc40.x86_64.img
}
Thank you again, hope that helps!
This is where it seems to be crashing:
Aug 08 22:24:50 fedora /usr/libexec/gdm-x-session[4236]: (EE)
Aug 08 22:24:50 fedora /usr/libexec/gdm-x-session[4236]: Fatal server error:
Aug 08 22:24:50 fedora /usr/libexec/gdm-x-session[4236]: (EE) Cannot open log file "/home/p520/.local/share/xorg/Xorg.pid-4236.log"
Aug 08 22:24:50 fedora /usr/libexec/gdm-x-session[4236]: (EE)
Aug 08 22:24:50 fedora /usr/libexec/gdm-x-session[4236]: (EE)
Aug 08 22:24:50 fedora /usr/libexec/gdm-x-session[4236]: Please consult the The X.Org Foundation support
Aug 08 22:24:50 fedora /usr/libexec/gdm-x-session[4236]: at http://wiki.x.org
Aug 08 22:24:50 fedora /usr/libexec/gdm-x-session[4236]: for help.
Aug 08 22:24:50 fedora /usr/libexec/gdm-x-session[4236]: (EE)
Aug 08 22:24:50 fedora /usr/libexec/gdm-x-session[4229]: Unable to run X server
Aug 08 22:24:50 fedora gdm-password][4112]: pam_unix(gdm-password:session): session closed for user p520
Aug 08 22:24:50 fedora audit[4112]: USER_END pid=4112 uid=0 auid=1000 ses=2 subj=system_u:system_r:xdm_t:s0-s0:c0.c1023 msg='op=PAM:session_close grantors=pam_selinux,pam_loginuid,pam_selinux,pam_keyinit,pam_namespace,pam_keyinit,pam_limits,pam_systemd,pam_unix,pam_gnome_keyring,pam_umask acct="p520" exe="/usr/libexec/gdm-session-worker" hostname=fedora addr=? terminal=/dev/tty2 res=success'
Aug 08 22:24:50 fedora audit[4112]: CRED_DISP pid=4112 uid=0 auid=1000 ses=2 subj=system_u:system_r:xdm_t:s0-s0:c0.c1023 msg='op=PAM:setcred grantors=pam_localuser,pam_unix,pam_gnome_keyring acct="p520" exe="/usr/libexec/gdm-session-worker" hostname=fedora addr=? terminal=/dev/tty2 res=success'
Aug 08 22:24:50 fedora gdm[1988]: Gdm: GdmDisplay: Session never registered, failing
Is there a reason you’re using X11 and not Wayland? Does the issue replicate under Wayland too?
This does not look normal
I do not have any of those installed on any of my systems.
Would you please post the output from sudo dmesg | grep -iE "nvidia|secure"
Well this will sound like a really stupid answer here, but honestly I don’t know. I honestly never bother changing, I just leave it to whatever it is. Generally I do know that most of my systems do run X11 for various reasons, usually low end systems, hahaha! So, I did think Fedora was running Wayland and maybe it should be changed to X11 as many other systems have success with it, but assumed it was still running Wayland. Perhaps another application randomly installed changed it? Everything else is expendable, the main purpose is getting DaVinci Resolve working again. I am sure I don’t need to tell you the direction Windows is going in, and I would like to avoid it as much as possible. If possible. So, do you know a way to change it pack to Wayland or better yet to test it and add a line to GRUB to change it for the session to test it out?
I am assuming it WAS initially running Wayland, seems I remember that when I set it up. But what do I know? Maybe it’s been X11 all along? Not sure. I type a lot, and am comfortable with the terminal command line, but just won’t know the correct commands. I can Google that though and see if there is a way to do it too. Thank you very much for all your help!
The graphical environment is session specific, and can be selected right after selecting the user from the login menu, using the cogwheel in the lower right corner.
Please also notice Jeff’s message above.
First I must apolgize, I am VERY tardy in replying - we are about to complete our move and so the months of August to October will be a bit crazy here. Anyway, I tried booting into Fedora again and clicked through all the things I could see but saw no graphical way to change from X11 to Wayland. Is there a way to do it via change to the GRUB boot line?
I don’t know have information about being able to provide GRUB with an argument to be passed to the kernel regarding which display server to load. However, the display server is usually selected from GDM, right after selecting the login user. If there is a setup in place to log in automatically, you can log out and log back in, but before entering the password, click on the gear wheel in the bottom right of the screen. What are the menu entries there?
Also, in order for those around here to be able to support you, it is recommended that you provide the requested information. Otherwise one has to make assumptions, which usually doesn’t help much. Please see the request in the post above, as your issues might have to do with the way nvidia drivers were installed on your system.
When logging in on the screen where you enter the password look at the gear in the lower right corner.
Greetings again! So a few things, first, when logging in, there is no gear icon in the lower right corner. The only things I can click on are the back arrow in front of the password box, the date calander on top in the center, and toward the right a man with various settings for … can’t think of the word but stuff like on screen keyboard, etc., and then to the right of that the button for internet, light or dark, etc.
Second, sorry for missing this earlier, the dmesg command output is:
https://paste.centos.org/view/raw/8917f274
I did notice one thing - “8.414075 nvidia: module verification failed: signature and/or required key missing:”. After updating nvidia I would assume that would take care of any such things, but obviously not. What did I do wrong?
Is there any more information I have forgotten to provide? I really appreciate all the help here, I really do. Thank you very much!
- Peter
The dmesg output you posted does not seem complete.
It should have entries earlier in the boot sequence, including a couple entries with the kernel command line and couple for secure boot.
Since the dmesg log is a rotating spool it is possible that it may have rotated out the earlier entries before the earliest time you show (6.970533 seconds). To get the full entry list for nvidia it may be necessary to reboot which should then allow including all nvidia related entries for the boot.
The message noted above is normal since that refers to the fact that it does not contain a fedora signature but is (or may be) signed by a locally generated key. Because it is not signed and distributed by fedora the message is provided.
Okay new link:
https://paste.centos.org/view/raw/77f353d2
To be honest, it looks the same as before. Makes sense in a way, both times the command was run just after a boot. So if there should be much more then - maybe that’s my problem, hahaha!! Is there a way to perhaps lengthen the spool to allow more entries? You did mention secure boot, not sure if you are referring to the linux kernel only or the system including Windows, but my PC has no secure boot enabled if that’s what you are referring to. If not, then forget I wrote anything.
Thank you very much again for your help and time, I’m sure there are lots of other things you could be doing instead. Thank you!
- Peter
How are you obtaining that and why are you posting it online instead of as part of the text here.
Please run that command sudo dmesg | grep -iE "nvidia|secure|command"
and then copy the text from your screen and paste it here if at all possible.
The issue seems not the spool size but that for some reason you are not getting the first part of the dmesg log that applies to nvidia & secure boot.
It should look something like this
# dmesg | grep -iE "nvidia|secure|command"
[ 0.000000] Command line: BOOT_IMAGE=(hd5,gpt2)/vmlinuz-6.10.4-200.fc40.x86_64 root=/dev/mapper/fedora_root-root ro rd.driver.blacklist=nouveau modprobe.blacklist=nouveau resume=UUID=1c1e5fa3-b2ae-4694-a13e-3364384da90d rd.lvm.lv=fedora_root/root rd.lvm.lv=fedora_root/swap rhgb quiet rd.driver.blacklist=nouveau modprobe.blacklist=nouveau
[ 0.000000] secureboot: Secure boot enabled
[ 0.000000] Kernel is locked down from EFI Secure Boot mode; see man kernel_lockdown.7
[ 0.003863] secureboot: Secure boot enabled
[ 0.038286] Kernel command line: BOOT_IMAGE=(hd5,gpt2)/vmlinuz-6.10.4-200.fc40.x86_64 root=/dev/mapper/fedora_root-root ro rd.driver.blacklist=nouveau modprobe.blacklist=nouveau resume=UUID=1c1e5fa3-b2ae-4694-a13e-3364384da90d rd.lvm.lv=fedora_root/root rd.lvm.lv=fedora_root/swap rhgb quiet rd.driver.blacklist=nouveau modprobe.blacklist=nouveau
[ 0.038375] Unknown kernel command line parameters "rhgb BOOT_IMAGE=(hd5,gpt2)/vmlinuz-6.10.4-200.fc40.x86_64", will be passed to user space.
[ 6.125978] input: HDA NVidia HDMI/DP,pcm=3 as /devices/pci0000:00/0000:00:03.1/0000:06:00.1/sound/card0/input20
[ 6.150894] input: HDA NVidia HDMI/DP,pcm=7 as /devices/pci0000:00/0000:00:03.1/0000:06:00.1/sound/card0/input21
[ 6.150964] input: HDA NVidia HDMI/DP,pcm=8 as /devices/pci0000:00/0000:00:03.1/0000:06:00.1/sound/card0/input22
[ 6.151019] input: HDA NVidia HDMI/DP,pcm=9 as /devices/pci0000:00/0000:00:03.1/0000:06:00.1/sound/card0/input23
[ 12.005176] nvidia: loading out-of-tree module taints kernel.
[ 12.129691] nvidia-nvlink: Nvlink Core is being initialized, major device number 234
[ 12.130427] nvidia 0000:06:00.0: vgaarb: VGA decodes changed: olddecodes=io+mem,decodes=none:owns=io+mem
[ 12.175220] NVRM: loading NVIDIA UNIX Open Kernel Module for x86_64 555.58.02 Release Build (akmods@eagle.home.domain) Thu Aug 15 14:55:59 CDT 2024
[ 12.572321] nvidia-modeset: Loading NVIDIA UNIX Open Kernel Mode Setting Driver for x86_64 555.58.02 Release Build (akmods@eagle.home.domain) Thu Aug 15 14:55:38 CDT 2024
[ 12.577027] [drm] [nvidia-drm] [GPU ID 0x00000600] Loading driver
[ 15.396247] [drm] Initialized nvidia-drm 0.0.0 20160202 for 0000:06:00.0 on minor 1
If necessary you can provide the command line and secure boot info with
mokutil --sb-state
and cat /proc/cmdline
Try `journalctl --no-hostname -b -g “nvidia|secure|command” – this should include the dmesg entries as well as the Xorg log entries.
Alright, so first, I am posting the links because … uh, I don’t know what else to do. Fedora is a hard install, not a virtual machine, so even for the paste links, I take a picture with my cellphone and come back upstairs to my other PC and copy it here from the picture. The outputs are usually a lot and in addition to being probably a bit lazy, I am also afraid I may make a typo which would throw off anyone nice enough to help me. Probably is a better way to do it but I don’t know it.
So here are some new links:
For sudo dmesg | grep -iE “nvidia|secure|command”
https://paste.centos.org/view/raw/17d489d8
For journalctl --no-hostname -b -g “nvidia|secure|command”
https://paste.centos.org/view/raw/26b26474
The other two commands are short enough to copy:
mokutil --sp-state
EFI variables are not supported on this system
cat /proc/cmdline
BOOT_IMAGE=/boot/vmlinuz-6.9.11-200.fc40.x86_64 root=/dev/nvme0n1p1
Hope that all helps! And thank you all again for your time and help!
- Peter