You are not logged in.
I got Nvidia GPU working last week with envycontrol, not it's not detected anymore. The only thing I did before it broke was full system upgrade.
journalctl -b | grep nvidia
May 08 06:14:15 archlinux kernel: nvidia: loading out-of-tree module taints kernel.
May 08 06:14:15 archlinux kernel: nvidia: module license 'NVIDIA' taints kernel.
May 08 06:14:15 archlinux kernel: nvidia: module verification failed: signature and/or required key missing - tainting kernel
May 08 06:14:15 archlinux kernel: nvidia: module license taints kernel.
May 08 06:14:15 archlinux kernel: nvidia-nvlink: Nvlink Core is being initialized, major device number 235
May 08 06:14:15 archlinux kernel: nvidia-nvlink: Unregistered Nvlink Core, major device number 235
May 08 06:14:16 archlinux systemd-modules-load[284]: Failed to insert module 'nvidia_uvm': No such device
May 08 06:14:18 archlinux kernel: nvidia-nvlink: Nvlink Core is being initialized, major device number 234
May 08 06:14:18 archlinux kernel: nvidia-nvlink: Unregistered Nvlink Core, major device number 234
May 08 06:14:19 archlinux kernel: nvidia-nvlink: Nvlink Core is being initialized, major device number 234
May 08 06:14:19 archlinux kernel: nvidia-nvlink: Unregistered Nvlink Core, major device number 234
May 08 06:14:19 archlinux systemd-modules-load[882]: Failed to insert module 'nvidia_uvm': No such device
May 08 06:14:19 archlinux systemd[1]: Started nvidia-powerd service.
May 08 06:14:19 archlinux nvidia-persistenced[1312]: Started (1312)
May 08 06:14:19 archlinux /usr/bin/nvidia-powerd[1310]: nvidia-powerd version:1.0(build 1)
May 08 06:14:19 archlinux nvidia-persistenced[1312]: Failed to query NVIDIA devices. Please ensure that the NVIDIA device files (/dev/nvidia*) exist, and that user 143 has read and write permissions for those files.
May 08 06:14:19 archlinux nvidia-persistenced[1312]: Shutdown (1312)
May 08 06:14:19 archlinux nvidia-persistenced[1309]: nvidia-persistenced failed to initialize. Check syslog for more details.
May 08 06:14:19 archlinux /usr/bin/nvidia-powerd[1310]: Allocate client failed 89
May 08 06:14:19 archlinux /usr/bin/nvidia-powerd[1310]: Failed to initialize RM Client
May 08 06:14:19 archlinux systemd[1]: nvidia-persistenced.service: Control process exited, code=exited, status=1/FAILURE
May 08 06:14:19 archlinux systemd[1]: nvidia-powerd.service: Main process exited, code=exited, status=1/FAILURE
May 08 06:14:19 archlinux systemd[1]: nvidia-powerd.service: Failed with result 'exit-code'.
May 08 06:14:19 archlinux systemd[1]: nvidia-persistenced.service: Failed with result 'exit-code'.
lspci -k | grep -A 2 -E "(VGA|3D)"
00:02.0 VGA compatible controller: Intel Corporation UHD Graphics 620 (rev 07)
Subsystem: Lenovo ThinkPad T480
Kernel driver in use: i915
dGPU was 00:01.0 but now it's undetected, so I think that it might be disabled after an upgrade but I'm not able to find a solution.
Offline
Please don't grep - post the complete system journal for the boot (a lot of relevant information will be under NVRM and any kind of bus error not covered by that either)
Offline
Here you go!
https://0x0.st/X8mO.txt
Offline
May 08 15:06:40 archlinux kernel: pci 0000:01:00.0: [10de:1d10] type 00 class 0x030200 PCIe Endpoint
May 08 15:06:40 archlinux kernel: pci 0000:01:00.0: BAR 0 [mem 0xec000000-0xecffffff]
May 08 15:06:40 archlinux kernel: pci 0000:01:00.0: BAR 1 [mem 0x90000000-0x9fffffff 64bit pref]
May 08 15:06:40 archlinux kernel: pci 0000:01:00.0: BAR 3 [mem 0xa0000000-0xa1ffffff 64bit pref]
May 08 15:06:40 archlinux kernel: pci 0000:01:00.0: BAR 5 [io 0xd000-0xd07f]
May 08 15:06:40 archlinux kernel: pci 0000:01:00.0: ROM [mem 0xfff80000-0xffffffff pref]
May 08 15:06:40 archlinux kernel: pci 0000:01:00.0: Enabling HDA controller
May 08 15:06:40 archlinux kernel: pci 0000:01:00.0: ROM [mem 0xfff80000-0xffffffff pref]: can't claim; no compatible bridge window
May 08 15:06:40 archlinux kernel: pci 0000:01:00.0: ROM [mem size 0x00080000 pref]: can't assign; no space
May 08 15:06:40 archlinux kernel: pci 0000:01:00.0: ROM [mem size 0x00080000 pref]: failed to assign
Did the update bring 6.8.9?
Does the GPU still show up w/ the LTS kernel?
Offline
I'm not sure what you meant in the first question. For the second one, it doesn't show up with LTS kernel in lspci as well.
Offline
The only thing I did before it broke was full system upgrade.
What packages got updated?
Does downgrading the kernel to the previous version bring back the GPU?
What if you add "pci=realloc" to the https://wiki.archlinux.org/title/Kernel_parameters ?
Offline
It was a full system upgrade, so everything should probably be latest version, I'm not sure which one of them got updated.
I tried downgrading to 6.8.7, same results.
Also adding "pci=realloc" to kernel parameter doesn't fix it.
Offline
Did you run a BIOS/UEFI update or change any of its settings or alter the hardware (add a PCI card or so)?
Offline
Sorry for bumping, but after upgrade i have similar issue:
==> ERROR: module not found: 'nvidia'
==> ERROR: module not found: 'nvidia_modeset'
==> ERROR: module not found: 'nvidia_uvm'
==> ERROR: module not found: 'nvidia_drm'
So it is something with the kernel.
Offline
That's on it's own not at all similar and this just suggests you either had a partial update or haven't rebuilt the nvidia drivers against the new kernel/gcc combination. If you're using an AUR package here make sure to rebuild them.
Offline
I always do this. This time i did this several times and always with the same error when nvidia-dkms is prepared.
BTW - full update.
==> ERROR: module not found: 'nvidia'
==> ERROR: module not found: 'nvidia_modeset'
==> ERROR: module not found: 'nvidia_uvm'
==> ERROR: module not found: 'nvidia_drm'
==> Generating module dependencies
==> Creating zstd-compressed initcpio image: '/boot/initramfs-linux-fallback.img'
==> WARNING: errors were encountered during the build. The image may not be complete.
error: command failed to execute correctly
[firekage@7500 6.8.9-arch1-2]$ yay -Ss nvidia-470xx | grep Installed
aur/nvidia-470xx-utils 470.239.06-1 (+92 1.04) (Installed)
aur/nvidia-470xx-dkms 470.239.06-1 (+92 1.04) (Installed)
[firekage@7500 6.8.9-arch1-2]$
The same error is reported when nvidia is uninstalled and only kernel is installed. Looks like there is no nvidia modules in kernel?
==> ERROR: module not found: 'nvidia'
==> ERROR: module not found: 'nvidia_modeset'
==> ERROR: module not found: 'nvidia_uvm'
==> ERROR: module not found: 'nvidia_drm'
==> Generating module dependencies
==> Creating zstd-compressed initcpio image: '/boot/initramfs-linux-fallback.img'
==> WARNING: errors were encountered during the build. The image may not be complete.
error: command failed to execute correctly
[firekage@7500 ~]$
I deleted nvidia. Updated again kernel. Rebooted. Installed again nvidia and the same. After reboot lsmod | grep nvidia shows nothing.
Edit: i did another thing. Removed nvidia, dkms. Rebooted. Installed nvidia not from the aur but from the pacman -S nvidia. The same. After rebot lsmod | grep nvidia shows nothing. It just does not work with the newest kernel.
Last edited by firekage (2024-05-08 19:06:07)
Offline
dkms status
and check the dkms build logs, but as V1del mentioned, this is a completely different situation from the OP
Edit: https://bbs.archlinux.org/viewtopic.php?id=295600
Last edited by seth (2024-05-08 20:06:40)
Offline
Did you run a BIOS/UEFI update or change any of its settings or alter the hardware (add a PCI card or so)?
I didn't, I did check if dGPU is disabled in BIOS, but it is not possible to do that on my specific laptop model.
Offline
Does
echo 1 | sudo tee /sys/bus/pci/devices/0000\:01\:00.0/remove
echo 1 | sudo tee /sys/bus/pci/rescan
bring up the device?
Also please post a journal for a pci=realloc" boot.
Possibly addressed by https://patchwork.kernel.org/project/li … edhat.com/
Offline
I just realized that after rebooting /proc/cmdline return to default, I followed section 1.9 in Kernel parameters of arch wiki. Before rebooting, I did check if the change was made, but it reverted back after reboot. I'm not sure if it's applied or not, if it was then here is the log.
https://0x0.st/X8QR.txt
Offline
pci=realloc is not in that journal.
This https://wiki.archlinux.org/title/Kernel … ng_cmdline is absolutely not what you want - follow the instructions for your bootloader.
If you're using UKI, see https://wiki.archlinux.org/title/Unifie … mmand_line
Offline
https://0x0.st/X8QF.txt
Is this correct?
Edit: Also,
echo 1 | sudo tee /sys/bus/pci/devices/0000\:01\:00.0/remove
echo 1 | sudo tee /sys/bus/pci/rescan
returns
tee: '/sys/bus/pci/devices/0000:01:00.0/remove': No such file or directory
1
1
Last edited by AungS8430 (2024-05-08 23:27:02)
Offline
May 09 05:29:44 archlinux kernel: Command line: initrd=\intel-ucode.img initrd=\initramfs-linux.img root=PARTUUID=6a5d5e3b-e959-4093-9cd5-cb6184e4613a zswap.enabled=0 rw rootfstype=ext4 pci=realloc
Yes, is.
But we stay at
May 09 05:29:44 archlinux kernel: pci 0000:01:00.0: ROM [mem size 0x00080000 pref]: can't assign; no space
May 09 05:29:44 archlinux kernel: pci 0000:01:00.0: ROM [mem size 0x00080000 pref]: failed to assign
Ok, you can't remove the device because it's not added itfp, but does the rescan do anything (in doubt give you more errors?
The only thing I did before it broke was full system upgrade.
Please post your pacman log.
Offline
Rescan doesn't return any errors.
Here's the pacman log.
https://0x0.st/X8Lm.log
Offline
Here're some excerpts reflectign your struggle to get the nvidia GPU working (I assume)
[2024-04-07T12:01:03+0700] [ALPM-SCRIPTLET] ==> ERROR: Missing dev kernel headers for module nvidia/550.67.
[2024-04-07T12:29:28+0700] [ALPM] removed nvidia-dkms (550.67-1)
[2024-04-07T12:29:28+0700] [ALPM] installed nvidia (550.67-5)
[2024-04-07T21:17:55+0700] [ALPM] installed nvidia-exec (0.1.2-1)
[2024-04-10T10:57:43+0700] [ALPM] removed nvidia-exec (0.1.2-1)
[2024-04-10T11:43:28+0700] [ALPM] installed nvidia-prime (1.0-4)
[2024-04-10T11:43:45+0700] [ALPM] installed nvidia-prime-rtd3pm (1.1-1)
At which point/date did the nvidia GPU no longer show up in lspci?
Also
[2024-04-01T07:03:54+0000] [ALPM] installed linux (6.8.2.arch2-1)
[2024-04-01T07:03:54+0000] [ALPM] installed linux-lts (6.6.23-1)
[2024-04-07T12:24:29+0700] [ALPM] upgraded linux (6.8.2.arch2-1 -> 6.8.4.arch1-1)
[2024-04-07T12:24:30+0700] [ALPM] upgraded linux-lts (6.6.23-1 -> 6.6.25-1)
are the only kernel installs/updates - how did you end up booting 6.8.7 ?
The rescan isn't supposed to generate errors but some output in the dmesg (about all the fancy PCI devices found) and - ideally - add the nvidia GPU?
Do you still have a journal from when the nvidia device showed up?
("sudo journalctl -b δ" allows you to go back in time the smaller the -negative- δ the older the boot, eg. try "-b -10")
Offline
Interesting - I have a work computer that uses Ubuntu 24.04 and this just started happening there. I can still use the GPU in hybrid mode when fully booted into the OS, but nvidia-persistenced is failing. Very strange. But maybe this isn't limited to Arch.
Offline
Do you have a journal from the ubuntu system?
Just persistenced failing isn't close to the situation the OP is encountering - the device cannot be addressed for them (ie. it doesn't even show up in lspci *at all*)
Offline
I don't have a system backup of 6.8.7 that has dGPU working. However, I have a backup of 6.8.2, which was when I did fresh arch installation. It was working on nonveau driver, will the log from 6.8.2 works?
Note: I tried to make PRIME, nvidia-exec, bumblebeed, some others in the wiki works but it doesn't or I don't like it so I ended up using envycontrol before the upgrade. I think that is potentially the issue, I messed up something.
Last edited by AungS8430 (2024-05-09 21:40:15)
Offline
You don't need backup of the system - "sudo journalctl -b δ # δ < 0" allows you to go back in time, eg. "sudo journalctl -b -10" will show the journal for 10 boots ago.
Since this the problem happens before any driver gets loaded, I'd be surprised if using nouveau would make any difference - but likewise if the older kernel works.
What if you just downgrade the kernel to 6.8.7 (you'll need nvidia-dkms or downgrade the nvidia and nvidia-utils packages along the kernel)
https://wiki.archlinux.org/title/Arch_Linux_Archive
Offline
I just went through boot journal saved, none of them have Nvidia GPU working. I did try to downgrade to 6.8.7 but didn't downgrade nvidia package or install nvidia-dkms, I think at least it should show up in lspci which it doesn't, should I try again but also install nvidia-dkms, will that make any difference?
Offline