You are not logged in.
Hello all. Recently had an issue after upgrading from linux-7.2.15.arch1-1 to linux-7.3.4.arch1-2 and nvidia-open-615.78.08-6 to nvidia-open-620.70.06-3. System would get past bootloader but would "hang" (primary monitor backlight only, secondary monitor dead).
Downgrading back to the previous versions remedied problem.
Mainly seeking to notify of potential issue, look for other affected users, and wait for a fix. Further information below.
System specifications:
- Motherboard: Gigabyte B650 GAMING X AX V2
- CPU: Ryzen 7 9700X
- GPU: NVIDIA GeForce RTX 2060 SUPER
- DE: KDE Plasma version 6.8.3
# journalctl -b -3 -k | grep nvidia
Nov 16 19:53:48 archlinux kernel: Command line: initrd=\initramfs-linux.img root=UUID=db65a1d3-b95e-45c9-8871-135d6697cadf rootflags=subvol=@ rw splash quiet nvidia.NVreg_PreserveVideoMemoryAllocations=1
Nov 16 19:53:48 archlinux kernel: Kernel command line: initrd=\initramfs-linux.img root=UUID=db65a1d3-b95e-45c9-8871-135d6697cadf rootflags=subvol=@ rw splash quiet nvidia.NVreg_PreserveVideoMemoryAllocations=1
Nov 16 19:53:55 maverick kernel: nvidia-gpu 0000:01:00.3: enabling device (0000 -> 0002)
Nov 16 19:53:55 maverick kernel: nvidia: loading out-of-tree module taints kernel.
Nov 16 19:53:55 maverick kernel: nvidia: module verification failed: signature and/or required key missing - tainting kernel
Nov 16 19:53:55 maverick kernel: nvidia-nvlink: Nvlink Core is being initialized, major device number 234
Nov 16 19:53:55 maverick kernel: nvidia 0000:01:00.0: vgaarb: VGA decodes changed: olddecodes=io+mem,decodes=none:owns=none
Nov 16 19:53:56 maverick kernel: nvidia-gpu 0000:01:00.3: i2c timeout error e0000000
Nov 16 19:53:56 maverick kernel: nvidia-modeset: Loading NVIDIA UNIX Open Kernel Mode Setting Driver for x86_64 620.70.06 Release Build (root@)
Nov 16 19:53:56 maverick kernel: [drm] [nvidia-drm] [GPU ID 0x00000100] Loading driver
Nov 16 19:53:56 maverick kernel: [drm] Initialized nvidia-drm 0.0.0 for 0000:01:00.0 on minor 1
Nov 16 19:53:56 maverick kernel: nvidia 0000:01:00.0: vgaarb: deactivate vga console
Nov 16 19:53:57 maverick kernel: fbcon: nvidia-drmdrmfb (fb0) is primary device
Nov 16 19:53:57 maverick kernel: nvidia 0000:01:00.0: [drm] fb0: nvidia-drmdrmfb frame buffer device
Nov 16 19:54:01 maverick kernel: nvidia-modeset: ERROR: GPU:0: Idling display engine timed out: 0x0000c57e:6:0:1209
Nov 16 19:54:03 maverick kernel: nvidia-modeset: ERROR: GPU:0: Idling display engine timed out: 0x0000c57e:6:0:1209
Nov 16 19:54:09 maverick kernel: nvidia-modeset: ERROR: GPU:0: Idling display engine timed out: 0x0000c57e:6:0:1209
Nov 16 19:54:18 maverick kernel: [drm:nv_drm_atomic_commit [nvidia_drm]] *ERROR* [nvidia-drm] [GPU ID 0x00000100] Failed to apply atomic modeset. Error code: -22Last 4 lines are particularly interesting.
Offline
"particularly interesting" … "grep nvidia" - nvidia lists most errors/debug info under NVRM, generally don't filter unless you know specifically what you're looking for.
Since you skipped a bunch of updates, it would likely be useful to know whether 620.70.06 exhibited this on 7.2.15.arch1-1 as well.
Online
I should have specified that I attempted nomodeset to no avail and was only able to get into my system after adding "3" to my kernel params. So I suspected a graphics issue after that.
My normal kernel params are as follows
root=UUID=db65a1d3-b95e-45c9-8871-135d6697cadf rootflags=subvol=@ rw splash quiet nvidia.NVreg_PreserveVideoMemoryAllocations=1And yes, I did try to remove "nvidia.NVreg_PreserveVideoMemoryAllocations=1". No, that didn't fix it.
As for 7.2.15.arch1-1 and 620.70.06, I observed the same behavior with the same four lines pertaining to modeset in the kernel logs (the ones I called "particularly interesting"). I did have to switch to nvidia-open-dkms, which isn't an exact match to nvidia-open in terms of the actual package, but should be close enough. Maybe I'm a fool for assuming that, though.
I am currently on linux-7.3.4.arch1-2 and nvidia-open-dkms-615.78.08-6 just fine, so the issue appears to be with 620.70.06 (both DKMS and non-DKMS, yes, I tried).
Here are some complete journalctl exports from problematic boots. I can provide more if necessary.
Boot with 7.2.15.arch1-1 and 620.70.06: https://privatebin.net/?df734bd868896c9 … YTLA4jReeP
Boot with 7.2.15.arch1-1 and 620.70.06 with nomodeset: https://privatebin.net/?2f919567fd58b5c … hsoUsUJLtJ
Edit: I should note that "nomodeset" did actually briefly show me plymouth whereas not having "nomodeset" did not. Probably inconsequential but worth noting anyways.
As for skipping "a bunch of updates"...not really. Originally I jumped from linux-7.2.15.arch1-1 to linux-7.3.4.arch1-2 (one update missed, linux-7.3.4.arch1-1), and from nvidia-open-615.78.08-6 to nvidia-open-620.70.06-3 (two updates missed, nvidia-open-620.70.06-1 and nvidia-open-620.70.06-2). At least those are the versions published on Gitlab. If missing that many updates on Arch is a crime, then I'm a worse criminal than that.
Last edited by chillsupthespine (Today 00:08:09)
Offline
So I suspected a graphics issue after that.
The point was that you're largely grepping for the wrong token.
nomodeset will not reasonably work w/ graphical targets (you're running on the simpledrm software emulation at best) but not even the multi-user.target is reached in either of the linked (or you possibly rebooted that w/ the power button, losing the tail of the journal)
Nov 17 18:10:55 maverick kernel: nvidia-gpu 0000:01:00.3: i2c timeout error e0000000
Nov 17 18:10:55 maverick kernel: ucsi_ccg 2-0008: i2c_transfer failed -110
Nov 17 18:10:55 maverick kernel: ucsi_ccg 2-0008: ucsi_ccg_init failed - -110
Nov 17 18:10:55 maverick kernel: ucsi_ccg 2-0008: probe with driver ucsi_ccg failed with error -110What happens when you blacklist ucsi_ccg ?
Online