You are not logged in.
Whenever my PC enters sleep, hibernate, or suspend, it fails to wake up properly. After attempting to resume, the display remains completely black. I don't get a login screen, desktop, TTY, cursor, or any visible output at all. Keyboard shortcuts such as switching to a TTY also do not appear to work. One thing I have noticed is that the keyboard backlight turns on when the system wakes, which makes me think the machine itself is resuming but the graphics/display stack is failing. Secure boot is disabled and im using KDE Plasma 6.6.5, discrete nvidia graphics rtx4060 595.71.05, kernel version 7.0.9-arch2-1.
Offline
Update: i was following the guide from this guy over here https://bbs.archlinux.org/viewtopic.php?id=313476 and my system is resuming normally when i switched it from s2idle[deep] to [s2idle]deep using the command sudo sh -c 'echo s2idle > /sys/power/mem_sleep' . Im gonna continue using this temporarily also i was doing some debugging and while on s2idle[deep] nvidia's resume hook was running and kde also woke up, but the monitor itself dint show anything. Pls do let me know what shall i do next. Im running wayland kde.
Offline
Do you have the same problem when only booting to and suspending from the mulit-user.target (2nd link below)
Do you have it w/ https://aur.archlinux.org/packages?O=0&K=580xx when disabling https://wiki.archlinux.org/title/NVIDIA … P_firmware (this does NOT work w/ the nvidia-open drivers)
keyboard backlight turns on when the system wakes, which makes me think the machine itself is resuming
Can you ssh into the system? Reboot it w/ https://wiki.archlinux.org/title/Keyboa … el_(SysRq) + REISUB (to maintain the journal which might explain what's going on)?
Offline
no i dont have https://aur.archlinux.org/packages?O=0&K=580xx and im using nvidia-open-dkms drivers. i tried doing reisub and it rebooted should i just paste the entirety of the file containing the journalctl here or you want a specific section?? Im sorry for the late reply i wasn't home.
Offline
No, the question was whether you have the problem when actually using those drivers (and disable the GSP)
Edit:
i tried doing reisub and it rebooted should i just paste the entirety of the file containing the journalctl here
sudo journalctl -b -1 | curl -s -H "Accept: application/json, */*" --upload-file - 'https://paste.c-net.org/'to upload the journal of the previous ("-1", -2, -3, … reach further back) boot to c-net, you get a link and key if you want to delete the post later.
Last edited by seth (2026-05-31 07:10:49)
Offline
I tried to ssh into the machine and it worked, I was able to run uptime and date command too. Also after i went into multi user I still wasn't able to wake up from suspend.
here is the file you wanted https://paste.c-net.org/ChancesOpposed
i haven't done the nvidia driver thing you mentioned yet but im using those nvidia-open-dkms drivers version 595.71.05.
Offline
May 30 20:40:44 <user> kernel: spd5118 2-0050: Failed to write b = 0: -6
May 30 20:40:44 <user> kernel: spd5118 2-0050: PM: dpm_run_callback(): spd5118_resume [spd5118] returns -6
May 30 20:40:44 <user> kernel: spd5118 2-0050: PM: failed to resume async: error -6That's a RAM temperature sensor.
Leaving aside the powerdevil spam and some BT hiccup there isn't much to explain this.
kwin starts to complain "Key repeat discarded, Wayland compositor doesn't seem to be processing events fast enough!" but only after "sysrq: Keyboard mode set to system default" so that doesn't count.
Also after i went into multi user I still wasn't able to wake up from suspend.
You used the sysrq, but did not isolate the multi-user.target before or after the suspend and there also seems no VT switch (did you at some point see any output from the REISUB?)
Next to that (2nd link below, no KDE) and the 580xx drivers: do you have a $5 mouse and keyboard from basic office supply? (there've been issues w/ some keyboards and udev events that broke VT switching but typically for X11)
Offline
I performed those tests after I recorded this journal report so they might not include the entry of them happening, In that journalctl i think i only did reisub before recording the report. No, Im currently using a razer mouse and a redragon keyboard they costed around 70 dollars combined. Also i dint see any output from reisub since my screen was completely black and it showed nothing i only knew reisub worked after it rebooted by alt+prtsc+b shorcut. I went into multiuser using the command sudo systemctl isolate multi-user.target followed by ctrl+alt+fn+f3 which worked and then i manually wrote systemctl suspend and it failed to wake up from there. then i tried the ssh and it worked.
Offline
No, Im currently using a razer mouse and a redragon keyboard they costed around 70 dollars combined.
Yup, the idea was to use something less fancy, more basic (the main offender is a keyboard w/ a bogus joystick device) - since the system wakes up (and the problem is the output) and there're no issues reported by the nvidia module (GSP or otherwise) we should not ignore that vector.
Offline
Hello Seth, I disconnected my external mouse and keyboard and rebooted and using my laptop's internal trackpad and keyboard I again tried it out and no luck it still went black with keyboard backlight functioning and the screen won't boot back up.
Offline
So test the 580xx drivers w/ GSP disabled.
Offline
Hey, Seth I installed the 580xx drivers and still nothing i did the usual systemctl suspend and it failed to wakeup from the black screen, I made sure to disable gsp too.
pacman -Qs nvidia
local/egl-gbm 1.1.3-1
The GBM EGL external platform library
local/egl-wayland 4:1.1.21-1
EGLStream-based Wayland external platform
local/egl-wayland2 1.0.1-1
EGLStream-based Wayland external platform (2)
local/egl-x11 1.0.5-1
NVIDIA XLib and XCB EGL Platform Library
local/libvdpau 1.5-4
Nvidia VDPAU library
local/libxnvctrl-580xx 580.159.04-1
NVIDIA NV-CONTROL X extension (580xx)
local/linux-firmware-nvidia 20260519-1
Firmware files for Linux - Firmware for NVIDIA GPUs and SoCs
local/nvidia-580xx-dkms 580.159.04-1
NVIDIA kernel modules - module sources (580xx)
local/nvidia-580xx-settings 580.159.04-1
Tool for configuring the NVIDIA graphics driver (580xx)
local/nvidia-580xx-settings-debug 580.159.04-1
Detached debugging symbols for nvidia-580xx-settings
local/nvidia-580xx-utils 580.159.04-1
NVIDIA drivers utilities (580xx)
local/vulkan-nouveau 1:26.1.1-2
Open-source Vulkan driver for Nvidia GPUs
local/xf86-video-nouveau 1.0.18-1 (xorg-drivers)
Open Source 3D acceleration driver for nVidia cards
Offline
![]()
Do you have a journal for that boot, too?
May 30 20:31:36 <user> kernel: DMI: LENOVO 83DV/LNVNB161216, BIOS NECN50WW 01/16/2026May 30 20:31:37 <user> kernel: ideapad_acpi VPC2004:00: Keyboard backlight control not available
May 30 20:31:40 <user> systemd[1]: Created slice Slice /system/systemd-backlight.
May 30 20:31:40 <user> systemd[1]: Starting Load/Save Screen Backlight Brightness of backlight:nvidia_0...
May 30 20:31:40 <user> systemd[1]: Finished Load/Save Screen Backlight Brightness of backlight:nvidia_0.
May 30 20:36:44 <user> systemd[1]: Created slice Slice /system/dbus-:1.2-org.kde.powerdevil.backlighthelper.
May 30 20:36:44 <user> systemd[1]: Started dbus-:1.2-org.kde.powerdevil.backlighthelper@0.service.
May 30 20:36:54 <user> systemd[1]: dbus-:1.2-org.kde.powerdevil.backlighthelper@0.service: Deactivated successfully.
May 30 20:38:15 <user> systemd[1]: Started dbus-:1.2-org.kde.powerdevil.backlighthelper@1.service.
May 30 20:38:26 <user> systemd[1]: dbus-:1.2-org.kde.powerdevil.backlighthelper@1.service: Deactivated successfully.That's a notebook? The only output (attached to the nvidia GPU) is the internal panel?
Can you attach some external monitor?
If you look *really* close at the panel when shining a flashlight onto it, can you see something on screen (except your mirror)?
Could be just the backlight…
Offline
Well, hey seth an external display is working and yes this is an notebook, i plugged the laptop to my tv and it seems to work but i cant see anything on the screen of the laptop even with the trick you told me to do.
Offline
On x11 check the output of "xrandr -q" (though I'd be surprised if the output was just disabled) and try to change the backlight level (figure whether brightnessctl works before the sleep and then assess the situation afterwards)
Transflection would not work if this is an OLED (and generally can be hard to make out since the panels aren't meant to do that)
Offline
Hey Seth, sorry for the late reply well im on wayland so I din't do the x11 test and my laptop monitor is ips, not oled. When I was checking the brightnessctl command they went through but the thing is the display was still completely black while there was output on tv, this was the output before i put the laptop to sleep
kscreen-doctor -o Output: 1 eDP-1 df8c061d-483c-4763-a10a-f269766fbd6f enabled connected priority 1 Panel replication source:0 Modes: 1:1920x1080@60.00! 2:1920x1080@144.00* Custom modes: None Geometry: 0,0 1920x1080 Scale: 1 Rotation: 1 Overscan: 0 Vrr: Never RgbRange: unknown HDR: incapable Wide Color Gamut: incapable ICC profile: none Color profile source: EDID Color power preference: prefer accuracy Brightness control: supported, set to 80% and dimming to 100% Color resolution: unknown Allow EDR: always Sharpness control: unsupported Automatic brightness: unsupported
and this was after i put it to sleep
sudo brightnessctl set 1% Updated device 'nvidia_0': Device 'nvidia_0' of class 'backlight': Current brightness: 1 (1%) Max brightness: 100 sudo brightnessctl set 100% Updated device 'nvidia_0': Device 'nvidia_0' of class 'backlight': Current brightness: 100 (100%) Max brightness: 100 [ kscreen-doctor -o Output: 1 eDP-1 df8c061d-483c-4763-a10a-f269766fbd6f enabled connected priority 2 Panel replication source:0 Modes: 1:640x480@59.94*! Custom modes: None Geometry: 0,0 640x480 Scale: 1 Rotation: 1 Overscan: 0 Vrr: Never RgbRange: unknown HDR: incapable Wide Color Gamut: incapable ICC profile: none Color profile source: EDID Color power preference: prefer accuracy Brightness control: supported, set to 80% and dimming to 100% Color resolution: unknown Allow EDR: always Sharpness control: unsupported Automatic brightness: unsupported Output: 2 HDMI-A-1 926a01ea-b22b-4f5c-9529-d96c22f127c0 enabled connected priority 1 HDMI replication source:0 Modes: 2:1920x1080@60.00*! 3:1920x1080@59.94 4:1920x1080@50.00 5:1920x1080@29.97 6:1920x1080@23.98 7:1680x1050@59.95 8:1280x1024@60.02 9:1152x864@75.00 10:1280x720@60.00 11:1280x720@59.94 12:1280x720@50.00 13:1280x720@29.97 14:1280x720@23.98 15:1024x768@60.00 16:800x600@60.32 17:720x576@50.00 18:720x480@59.94 19:640x480@59.94 20:640x480@59.93 Custom modes: None Geometry: 0,0 1920x1080 Scale: 1 Rotation: 1 Overscan: 0 Vrr: incapable RgbRange: unknown HDR: incapable Wide Color Gamut: incapable ICC profile: none Color profile source: sRGB Color power preference: prefer efficiency and performance Brightness control: supported, set to 100% and dimming to 100% Color resolution: unknown Allow EDR: unsupported Sharpness control: unsupported Automatic brightness: unsupported
my internal display monitor is 1080p 144hz panel.
Offline
this was the output before i put the laptop to sleep
The situation *after* the sleep when the display is dysfunctional would be more interesting.
Please use [code][/code] tags and maintain line breaks for readability.
Though assuming the second kscreen-doctor output stashed in there is actually significant:
Output: 1 eDP-1 df8c061d-483c-4763-a10a-f269766fbd6f enabled connected priority 2 Panel replication source:0 Modes: 1:640x480@59.94*!The panel is at VGA, you might have lost the EDID?
What if you copy it from the working state and statically inject it from the OS?
https://wiki.archlinux.org/title/Kernel … s_and_EDID
Offline
Seth i tried this and it just won't wake up T_T T_T.
Offline
The status quo ante was that the system woke up but the eDP was dysfunctional (while you could still use SSH or the HDMI output)?
Is this still the case?
Please post your complete system journal for the boot after the botched wakeup (where the EDID is supposed to be injected by the OS):
sudo journalctl -b | curl -s -H "Accept: application/json, */*" --upload-file - 'https://paste.c-net.org/'Offline
Hmm Seth, this is the result of kscreen-doctor -o after the screen went black.
Output: 1 eDP-1 df8c061d-483c-4763-a10a-f269766fbd6f
enabled
connected
priority 2
Panel
replication source:0
Modes: 1:1920x1080@144.00*! 2:1920x1080@60.00
Custom modes: None
Geometry: 0,0 1920x1080
Scale: 1
Rotation: 1
Overscan: 0
Vrr: Never
RgbRange: unknown
HDR: incapable
Wide Color Gamut: incapable
ICC profile: none
Color profile source: EDID
Color power preference: prefer accuracy
Brightness control: supported, set to 80% and dimming to 100%
Color resolution: unknown
Allow EDR: always
Sharpness control: unsupported
Automatic brightness: unsupported
Output: 2 HDMI-A-1 926a01ea-b22b-4f5c-9529-d96c22f127c0
enabled
connected
priority 1
HDMI
replication source:0
Modes: 3:1920x1080@60.00*! 4:1920x1080@59.94 5:1920x1080@50.00 6:1920x1080@29.97 7:1920x1080@23.98 8:1680x1050@59.95 9:1280x1024@60.02 10:1152x864@75.00 11:1280x720@60.00 12:1280x720@59.94 13:1280x720@50.00 14:1280x720@29.97 15:1280x720@23.98 16:1024x768@60.00 17:800x600@60.32 18:720x576@50.00 19:720x480@59.94 20:640x480@59.94 21:640x480@59.93
Custom modes: None
Geometry: 0,0 1920x1080
Scale: 1
Rotation: 1
Overscan: 0
Vrr: incapable
RgbRange: unknown
HDR: incapable
Wide Color Gamut: incapable
ICC profile: none
Color profile source: sRGB
Color power preference: prefer efficiency and performance
Brightness control: supported, set to 100% and dimming to 100%
Color resolution: unknown
Allow EDR: unsupported
Sharpness control: unsupported
Automatic brightness: unsupportedand here is the journalctl you requested "https://paste.c-net.org/PlasmaPupil"
Offline
kscreen considers the output to be there, enabled and running at 1080p144Hz - eDP and HDMI are covering the same area, so we don't expect any empty space.
There's a lot of powerdevil about the HDMI output but no record of the eDP one ![]()
1. https://wiki.archlinux.org/title/KDE#Va … n_problems
2.
Jun 14 01:33:13 <user> org_kde_powerdevil[1459]: [ 2550] Open failed for /dev/i2c-3, errno=EACCES(-13): Permission denied in file i2c_bus_core.c near line 433
Jun 14 01:33:13 <user> org_kde_powerdevil[1459]: [ 2550] Checking for group i2c...
Jun 14 01:33:13 <user> org_kde_powerdevil[1459]: [ 2550] Group i2c exists
Jun 14 01:33:13 <user> org_kde_powerdevil[1459]: [ 2550] WARNING: Current user <user> is NOT a member of group i2c=> add yourself to the i2c group and re-login.
Offline
Hey Seth, I did as you said but to no avail the screen din't return from its complete black state.
Offline
Hey Seth, what should I do? any updates??
Offline
I had your thread still in a background tab and am waiting for my brain to come up w/ something… ![]()
Just ftr, you've successfully disabled POWERDEVIL_NO_DDCUTIL=1 and added yourself to the i2c group and powerdevil is no longer trying to control any outputs nor does it issue the i2c errors?
To rule out KDE, you're losing the eDP when only booting the multi-user.target and sleeping from there (never started any GUI stuff) and have the EDID injected into the system?
Offline
Hey Seth, I think I found something I changed my graphics from discrete to dynamic from the boot menu and it is restoring from sleep (I literally can't believe that was the solution all along T_T omg) tho Seth im getting 5-6 messages of spd5118_resume failed during the process of booting from sleep.
Offline