Screen goes black a short period after booting

isopropylpod · 2025-01-03 04:06:47

This is an issue I've had for quite some time. I got a 4K monitor awhile back, and it has been causing a lot of issues for me. The main is that when I turn on my PC, unless I turn it off and turn it on at just the right time, it will say that it cannot find input, which wouldn't be something to bring up here, except that unless it manages to turn on, my other display (which has worked flawlessly for nearly a decade now) will shut down around 10 seconds later, taking the ability to input anything with it. The weirdest thing about this whole issue is that the issue persists if I leave the monitor off, and the issue even persists if I completely unplug the 4K monitor from my GPU. This has persisted through two complete re-installs, and has affected both Hyprland and KDE Plasma. I have an Nvidia GPU and I am using the 'Nvidia' proprietary driver. I have listed the past three journals, with 1 being 2 boots ago, 2 being 1 boot ago, and 3 being the current boot that I was able to get working.

https://gist.github.com/Isolyth/f4c854d … 6ff10d6d3c

isopropylpod · 2025-01-03 06:04:30

It has now become impossible to log in to the system regardless of timing. I left to do something for a few minutes and I guess I missed a sleep setting or something and when I came back my displays were both black (not off or saying no input, just displaying solid black) and now when I boot up the system I can get into SDDM with proper timing but once I click log in both displays go dark with no recovery.

Edit: When I plug in the problematic monitor to my motherboard (IGPU) and leave my other plugged into my GPU, I am able to boot normally and the monitor detects the port, but the GPU plugged in monitor goes dark and the 4K display freezes completely (though continues to show the last image), and when I have just the 4K display plugged in things seem to be okay, hate to say it but it could be a GPU hardware issue? It's so weird because this only started happening with this display, how could I confirm it?

Edit 2: I tested with x11 and it worked for one boot, but then I set my 4k display's refreshrste to 144hz and the system now black screens on x11 as well

Last edited by isopropylpod (2025-01-03 08:14:46)

isopropylpod · 2025-01-03 09:08:25

Here is an updated journal
https://gist.github.com/Isolyth/a2b77c7 … 2a1b776bc8

I have switched to an LTS kernel (the kernel bug listed in the journal also happens identically for the base kernel as well, switching did not fix it) and Nvidia-open, which also didn't resolve the issue. In this specific launch, my 4K display did not turn on, my 1440p display was on for about a minute, flickered twice, then went dark indefinitely about 30 seconds later. Each run is random on if SDDM will show on both screens or just my smaller one, and if X or Wayland will show on both screens (a situation which I haven't gotten in Wayland yet) or just the smaller one, and how long the small display lasts.

seth · 2025-01-03 16:28:29

Please don't bump edit your previous boot to mend it if nobody has yet replied.
You've created the illusion of ongoing activity in this thread.

You've two VGA devices, an nvidia and an AMD one

Jan 03 03:30:24 theseus kernel: amdgpu 0000:0d:00.0: [drm] Cannot find any crtc or sizes

There're no outputs on the AMD one, you're

Jan 03 03:30:23 theseus kernel: simple-framebuffer simple-framebuffer.0: [drm] Registered 1 planes with drm panic

loading the simpledrm driver and

Jan 03 03:30:23 theseus kernel: NVRM: loading NVIDIA UNIX x86_64 Kernel Module  565.77  Wed Nov 27 23:33:08 UTC 2024
Jan 03 03:32:04 theseus kernel: NVRM: GPU at PCI:0000:01:00: GPU-df4f7d44-3abc-3e01-e3cd-ca43d6110fd2
Jan 03 03:32:04 theseus kernel: NVRM: Xid (PCI:0000:01:00): 119, pid=1332, name=kitty, Timeout after 6s of waiting for RPC response from GPU0 GSP! Expected function 10 (FREE) (0xcaf00003 0x0).

teh nvidia one has some serious trouble.

Add "nvidia.NVreg_EnableGpuFirmware=0 nvidia_drm.modeset=1" to the https://wiki.archlinux.org/title/Kernel_parameters which will hopefully fix your problem by avoiding the GSP (not required for the non-open driver) and blocking the simpledrm device.

isopropylpod · 2025-01-03 19:36:42

Sorry for the edits, was just trying to clarify the situation, anyway, I added that parameter but the system is still unable to boot into wayland at all, and X still has issues on random boots. Here are the journals, labeled wayland and X.

https://gist.github.com/Isolyth/52602ef … 2af9437d12

seth · 2025-01-03 20:28:40

Jan 03 14:30:47 theseus kernel: NVRM: GPU at PCI:0000:01:00: GPU-df4f7d44-3abc-3e01-e3cd-ca43d6110fd2
Jan 03 14:30:47 theseus kernel: NVRM: Xid (PCI:0000:01:00): 62, pid='<unknown>', name=<unknown>, 5218827c c09a060c 00000000 202c724e 202c63b6 202c5bbe 202c489c 202c4cea
Jan 03 14:30:51 theseus kernel: NVRM: Xid (PCI:0000:01:00): 154, pid='<unknown>', name=<unknown>, GPU recovery action changed from 0x0 (None) to 0x1 (GPU Reset Required)

Jan 03 14:31:47 theseus kernel: NVRM: loading NVIDIA UNIX x86_64 Kernel Module  565.77  Wed Nov 27 23:33:08 UTC 2024
Jan 03 14:32:35 theseus kernel: NVRM: GPU at PCI:0000:01:00: GPU-df4f7d44-3abc-3e01-e3cd-ca43d6110fd2
Jan 03 14:32:35 theseus kernel: NVRM: Xid (PCI:0000:01:00): 62, pid='<unknown>', name=<unknown>, 5218827c c09a060c 00000000 202c724e 202c63b6 202c5bbe 202c489c 202c4cea

62 is "Internal micro-controller halt (newer drivers)" and either a hardware or driver issue, 154 isn't documented.
https://docs.nvidia.com/deploy/xid-errors/index.html

But both journals show SDDM starting and you logging into plasma.
Are those auto-logins or do you actually get to see the DM and then things only go south once you log into plasma?

The X11 one has

Jan 03 14:32:20 theseus kernel: nvidia-modeset: WARNING: GPU:0: HDMI FRL link training failed.

That's your 4k output.

1. cable: 4k@144Hz will require HDMI2.1 and an "Ultra High Speed" cable, 4k@60Hz you'll get away w/ a "Premium High Speed" cable (and yes, that bullshit is the official grading system)
2.

Jan 03 14:31:47 theseus kernel: DMI: ASRock B650I Lightning WiFi/B650I Lightning WiFi, BIOS 2.02.AS01 11/17/2023

Is the GPU sufficiently powered (PEG slot and dedicated 6/8-pin power supply)? For multiscreen setups, the nvidia driver will run in a higher power mode.
3. for KDE/kscreen specifically see https://wiki.archlinux.org/title/KDE#Un … lution_set and wipe ~/.local/share/kscreen/ - especially when switching output connections (between GPU and APU)

Also please post your Xorg log, https://wiki.archlinux.org/title/Xorg#General

isopropylpod · 2025-01-04 00:45:59

I am able to get into SDDM sometimes; occasionally only my smaller monitor will get in and the other will remain dark, this typically spells doom once I log in, other times both will be on, which generally allows me to get into X (not Wayland though)
1) I have a sufficient cable, I got it when I got this display and I have been able to use it in 4k@144 before, back when the turn it off and back on during boot trick was working
2) it should, again, it worked before, and when I do get to X btop shows it pulling 100~ watts
3) I will attempt clearing that out, hopefully that could fix it crashing (sometimes with the screen strobing rapidly) in SDDM

I could not find a Xorg folder in either location mentioned by the wiki

Clearing the kscreen folder did not get me into SDDM for any functional amount of time, small screen shut down a few seconds after getting in to it, which is the same result as before. I'm going to attempt reinstalling arch without ever connecting the 4K display to see if that helps - if not, I have no other explanation here other than a hardware failure.

I am pretty certain now this is a hardware failure on my GPU; doing a full reinstall causes my smaller monitor (the only one connected) to go dark after about 5 seconds in SDDM. I don't know of anyway to get my journal off the machine without having some sort of GUI to upload it, but I see the same error block as before, with the messages (in order, plus RPC history and such) being
... Timeout after 6s of waiting for RPC response from GPU0 GSP! Expected function 76 (GSP_RM_CONTROL) (0X20800a70)
Nvidia-modeset: ERROR: GPU:0: Failed to query display engine channel state: 0x0000c76e:6:0:0x00000065
After applying the parameters to disable the GSP, I get the same error as I did before regarding the GPU recovery action, with a new mention of CTX SWITCH TIMEOUT, error codes seem to be 62, 154, 56, 56, 56, 56, 109, in that order, SDDM then reports the context being lost and the graphics device being lost, then NVRM reports 56, 56, 56, 56, 16, 109

It still seems to be possible to boot by turning off the monitor and timing turning it back on correctly, as I am able to boot into Hyprland (My old install) by doing this. Here is the journal from that attempt. Is it possible there is an issue when SDDM shuts down for KDE to start? https://gist.github.com/Isolyth/3f1e776 … e91ce5d8fd

Last edited by isopropylpod (2025-01-04 07:01:34)

seth · 2025-01-04 17:42:08

I could not find a Xorg folder in either location mentioned by the wiki

For sddm it'll be /var/log/Xorg.0.log - no whatsoever "xorg folder".

Clearing the kscreen folder did not get me into SDDM

The kscreen config of your user session has no impact on the DM

reinstalling arch without ever connecting the 4K display to see if that helps

…won't help /spoiler

go dark after about 5 seconds in SDDM

Could simply be dpms?

I don't know of anyway to get my journal off the machine without having some sort of GUI to upload it

1st link below

sudo journalctl -b | curl -F 'file=@-' 0x0.st

boot by turning off the monitor and timing turning it back on correctly, as I am able to boot into Hyprland

You do not "boot" into Hyprland - not even SDDM.
Whatever your DM and esp. your GUI session do happens *way* after the boot.
If you can utilize the system w/o GUI (you've one or both outputs here) this is unlikely to be a hardware problem.

Timeout after 6s of waiting for RPC response from GPU0 GSP! Expected function 76 (GSP_RM_CONTROL) (0X20800a70)
Nvidia-modeset: ERROR: GPU:0: Failed to query display engine channel state: 0x0000c76e:6:0:0x00000065

"nvidia.NVreg_EnableGpuFirmware=0 nvidia_drm.modeset=1" … but there's nothing like that in the posted journal.

Can you deactivate the AMD gpu in the firmware or add "module_blacklist=amdgpu" to prevent the GUI from running on the "wrong" GPU?

isopropylpod · 2025-01-05 00:12:21

Here is the journal with the parameters set, the AMD gpu disabled via the module blacklist, seemingly everything I need to do. SDDM starts fine, but once I log in I just get a black screen. It appears from my limited research that XID 45 is hardware failure, based on https://www.reddit.com/r/pop_os/comment … ermittent/ and https://forums.developer.nvidia.com/t/x … ues/271392 , in the latter the hardware failure status is doubted, but given a lack of a resolution and it being only on a single of multiple GPUs they own (if I read it right), I am fairly certain now this is a hardware failure (though why SDDM -> KDE triggers it, and Hyprland doesn't, I don't know)

https://gist.github.com/Isolyth/9bc7733 … d48e20dce3

Arch Linux

#1 2025-01-03 04:06:47

Screen goes black a short period after booting

#2 2025-01-03 06:04:30

Re: Screen goes black a short period after booting

#3 2025-01-03 09:08:25

Re: Screen goes black a short period after booting

#4 2025-01-03 16:28:29

Re: Screen goes black a short period after booting

#5 2025-01-03 19:36:42

Re: Screen goes black a short period after booting

#6 2025-01-03 20:28:40

Re: Screen goes black a short period after booting

#7 2025-01-04 00:45:59

Re: Screen goes black a short period after booting

#8 2025-01-04 17:42:08

Re: Screen goes black a short period after booting

#9 2025-01-05 00:12:21

Re: Screen goes black a short period after booting

Board footer