Black screen after resuming from sleep

cstdio · 2024-10-09 20:58:10

On an nvme. Everything is on a 2T nvme.

Device            Start        End    Sectors  Size Type
/dev/nvme0n1p1     2048    2099199    2097152    1G EFI System
/dev/nvme0n1p2  2099200   10487807    8388608    4G Linux swap
/dev/nvme0n1p3 10487808 4000796671 3990308864  1.9T Linux root (x86-64)

Last edited by cstdio (2024-10-09 21:17:04)

gerwazy · 2024-10-09 21:11:13

OS: Arch Linux x86_64
Host: B550 GAMING X V2 (-CF)
Kernel: Linux 6.11.2-arch1-1
Shell: zsh 5.9
DE: KDE Plasma 6.1.5
WM: KWin (X11)
WM Theme: summaculate-night-blur
Theme: Breeze (Dark) [Qt], Breeze-Dark [GTK2], Breeze [GTK3]
Icons: breeze-dark [Qt], breeze-dark [GTK2/3/4]
Font: Noto Sans (10pt) [Qt], Noto Sans (10pt) [GTK2/3/4]
Cursor: breeze (24px)
Terminal: kitty 0.36.4
CPU: AMD Ryzen 7 5800X (16) @ 4.85 GHz - 44.2°C
GPU: NVIDIA GeForce GTX 1650 - 51.0°C [Discrete]
Memory: 8.54 GiB / 62.72 GiB (14%)
Swap: Disabled
Disk (/): 30.40 GiB / 48.91 GiB (62%) - ext4
Disk (/home): 400.83 GiB / 1.67 TiB (23%) - ext4

sda 8:0 0 476,9G 0 disk
├─sda1 8:1 0 25G 0 part
├─sda2 8:2 0 25G 0 part
├─sda3 8:3 0 25G 0 part
├─sda4 8:4 0 1K 0 part
├─sda5 8:5 0 12G 0 part
└─sda6 8:6 0 389,9G 0 part
nvme0n1 259:0 0 1,9T 0 disk
├─nvme0n1p1 259:1 0 2G 0 part /boot/efi
├─nvme0n1p2 259:2 0 50G 0 part
├─nvme0n1p3 259:3 0 50G 0 part /
├─nvme0n1p4 259:4 0 63G 0 part
├─nvme0n1p5 259:5 0 1,7T 0 part /home
└─nvme0n1p6 259:6 0 2G 0 part

seth · 2024-10-09 21:25:20

Sorry, the question was directed at Gasp0de - if they're running from an SSD, we can close the nvme vector.

herbiejhopkins · 2024-10-10 17:28:36

I've also been having the same issue since I upgraded to 6.11. System used to suspend just fine on 6.10 but now hard locks when trying to resume. Last thing in my journalctl output is similar to others:

Oct 10 13:56:52 arcturus kernel: PM: suspend entry (deep)
Oct 10 13:56:52 arcturus kernel: Filesystems sync: 0.058 seconds

I'm also on an AMD system with single nvme drive.

Operating System: Arch Linux
KDE Plasma Version: 6.1.5
KDE Frameworks Version: 6.6.0
Qt Version: 6.7.3
Kernel Version: 6.11.2-arch1-1 (64-bit)
Graphics Platform: Wayland
Processors: 24 × AMD Ryzen 9 7900 12-Core Processor
Memory: 61.9 GiB of RAM
Graphics Processor: AMD Radeon RX 7900 XT
Manufacturer: ASUS
Product Name: B650E-I

Reuter · 2024-10-11 12:54:35

Gasp0de wrote:

Just chiming in here to say that I have the same issues with the 6.11 Linux kernel, I'm on a Thinkpad laptop, AMD Ryzen processor and integrated graphics, no Nvidia anywhere in or on my system, and I am using Gnome.

Same here. Thinkpad AMD, but Plasma instead of Gnome.

Sometimes the device wakes up to a black screen, sometimes to the SDDM Login screen. It's noticable that the clock on said login screen has been frozen way before the time of the attempted wakeup.

Sp4rks · 2024-10-12 08:47:49

My AMD desktop also had sleep/resume issues after upgrading to 6.11.

Similar to OP, USB devices turned on fine, but screen resumed with only a stuck mouse cursor. journalctl output didn't show anything that's not already been posted/discussed.
My machine does have 3 NVME drives for storage, but a Kingston SSD is where my Arch Linux resides.

OS: Arch Linux x86_64
Host: B650 GAMING X AX
Resolution: 1920x1080
DE: Plasma 6.2.0
WM: kwin
CPU: AMD Ryzen 7 7800X3D (16) @ 5.0
GPU: AMD ATI Radeon RX 7900 XT/7900
GPU: AMD ATI 13:00.0 Raphael
Memory: 9507MiB / 31046MiB

sda 8:0 0 894.3G 0 disk
├─sda1 8:1 0 512M 0 part /boot
├─sda2 8:2 0 55.9G 0 part /
└─sda3 8:3 0 834.6G 0 part /home

If anyone reading this wants a quick easy fix... use the linux-lts kernel for now. Sleep/resume worked fine again.

seth · 2024-10-12 15:19:48

My machine does have 3 NVME drives for storage, but a Kingston SSD is where my Arch Linux resides.

If the nvme aren't really used (mounted), that would completely rule out that problem.
=> https://gitlab.freedesktop.org/drm/amd/-/issues
As you can see, there's a good chunk of 6.11 related issues already: https://gitlab.freedesktop.org/drm/amd/ … ge_size=20

One in particular looks interesting, try

amdgpu.aspm=0 amdgpu.runpm=0 amdgpu.bapm=0 pcie_aspm=off amdgpu.dpm=0

but dpm=0 in the past has lead to boot failures…

gerwazy · 2024-10-12 17:08:08

It seems there was a change, likely after today's update of nvidia-utils or xorg-wayland. Now, the computer no longer goes into sleep mode; instead, only the screen turns off.

Last edited by gerwazy (2024-10-12 17:12:14)

seth · 2024-10-12 18:43:35

Do you have a journal covering such failed sleep attempt?
After the screen turns off, can you wake it and continue to use the system?

gerwazy · 2024-10-12 19:33:34

After the screen turns off, when the system does not enter sleep mode, I can still wake it up to resume normal operation.

journalctl -b -1 -p err
paź 12 18:07:47 Linux kernel: 
paź 12 18:08:04 Linux pipewire[720]: spa.v4l2: '/dev/video0' VIDIOC_QUERYCTRL: Błąd wejścia/wyjścia
paź 12 18:08:04 Linux pipewire[720]: spa.v4l2: '/dev/video0' VIDIOC_QUERYCTRL: Błąd wejścia/wyjścia
paź 12 18:08:05 Linux systemd-coredump[953]: [?] Process 949 (smartctl) of user 0 dumpe
d core.
                                              
                                              Stack trace of thread 949:
                                              #0  0x00007f18f02a53f4 n/a (libc.so.6 + 0x963f4)
                                              #1  0x00007f18f024c120 raise (libc.so.6 + 0x3d120)
                                              #2  0x00007f18f02334c3 abort (libc.so.6 + 0x244c3)
                                              #3  0x00007f18f0234354 n/a (libc.so.6 + 0x25354)
                                              #4  0x00007f18f02af765 n/a (libc.so.6 + 0xa0765)
                                              #5  0x00007f18f02b1c4c n/a (libc.so.6 + 0xa2c4c)
                                              #6  0x00007f18f02b45ce __libc_free (libc.so.6 + 0xa55ce)
                                              #7  0x000064c6246588d2 n/a (smartctl + 0x448d2)
                                              #8  0x00007f18f024e891 n/a (libc.so.6 + 0x3f891)
                                              #9  0x00007f18f024e95e exit (libc.so.6 + 0x3f95e)
                                              #10 0x00007f18f0234e0f n/a (libc.so.6 + 0x25e0f)
                                              #11 0x00007f18f0234ecc __libc_start_main (libc.so.6 + 0x25ecc)
                                              #12 0x000064c624620af5 n/a (smartctl + 0xcaf5)
                                              ELF object binary architecture: AMD x86-64
paź 12 18:23:08 Linux kernel: nvidia 0000:04:00.0: PM: pci_pm_suspend(): nv_pmops_suspend [nvidia] returns
 -5
paź 12 18:23:08 Linux kernel: nvidia 0000:04:00.0: PM: dpm_run_callback(): pci_pm_suspend returns -5
paź 12 18:23:08 Linux kernel: nvidia 0000:04:00.0: PM: failed to suspend async: error -5
paź 12 18:23:08 Linux kernel: PM: Some devices failed to suspend, or early wake event detected
paź 12 18:23:09 Linux kernel: nvidia 0000:04:00.0: PM: pci_pm_suspend(): nv_pmops_suspend [nvidia] returns
 -5
paź 12 18:23:09 Linux kernel: nvidia 0000:04:00.0: PM: dpm_run_callback(): pci_pm_suspend returns -5
paź 12 18:23:09 Linux kernel: nvidia 0000:04:00.0: PM: failed to suspend async: error -5
paź 12 18:23:09 Linux kernel: PM: Some devices failed to suspend, or early wake event detected
paź 12 18:23:09 Linux systemd-sleep[1289]: Failed to put system to sleep. System resumed again: Input/outp
ut error
paź 12 18:23:09 Linux systemd[1]: Failed to start System Suspend.
paź 12 18:38:15 Linux kernel: nvidia 0000:04:00.0: PM: pci_pm_suspend(): nv_pmops_suspend [nvidia] returns
 -5
paź 12 18:38:15 Linux kernel: nvidia 0000:04:00.0: PM: dpm_run_callback(): pci_pm_suspend returns -5
paź 12 18:38:15 Linux kernel: nvidia 0000:04:00.0: PM: failed to suspend async: error -5
paź 12 18:38:15 Linux kernel: PM: Some devices failed to suspend, or early wake event detected
paź 12 18:38:15 Linux kernel: nvidia 0000:04:00.0: PM: pci_pm_suspend(): nv_pmops_suspend [nvidia] returns
 -5
paź 12 18:38:15 Linux kernel: nvidia 0000:04:00.0: PM: dpm_run_callback(): pci_pm_suspend returns -5
paź 12 18:38:15 Linux kernel: nvidia 0000:04:00.0: PM: failed to suspend async: error -5
paź 12 18:38:15 Linux kernel: PM: Some devices failed to suspend, or early wake event detected
paź 12 18:38:15 Linux systemd-sleep[1476]: Failed to put system to sleep. System resumed again: Input/outp
ut error
paź 12 18:38:15 Linux systemd[1]: Failed to start System Suspend.
paź 12 18:53:16 Linux kernel: nvidia 0000:04:00.0: PM: pci_pm_suspend(): nv_pmops_suspend [nvidia] returns
 -5
paź 12 18:53:16 Linux kernel: nvidia 0000:04:00.0: PM: dpm_run_callback(): pci_pm_suspend returns -5
paź 12 18:53:16 Linux kernel: nvidia 0000:04:00.0: PM: failed to suspend async: error -5
paź 12 18:53:16 Linux kernel: PM: Some devices failed to suspend, or early wake event detected
paź 12 18:53:17 Linux kernel: nvidia 0000:04:00.0: PM: pci_pm_suspend(): nv_pmops_suspend [nvidia] returns
 -5
paź 12 18:53:17 Linux kernel: nvidia 0000:04:00.0: PM: dpm_run_callback(): pci_pm_suspend returns -5
paź 12 18:53:17 Linux kernel: nvidia 0000:04:00.0: PM: failed to suspend async: error -5
paź 12 18:53:17 Linux kernel: PM: Some devices failed to suspend, or early wake event detected
paź 12 18:53:17 Linux systemd-sleep[1650]: Failed to put system to sleep. System resumed again: Input/outp
ut error--
paź 12 18:53:17 Linux systemd[1]: Failed to start System Suspend.
paź 12 19:27:31 Linux kwin_x11[6265]: This application failed to start because no Qt platform plugin could
 be initialized. Reinstalling the application may fix this problem.
                                       
                                       Available platform plugins are: xcb, linuxfb, minimalegl, eglfs, vkkhrdisplay, wayl
and-egl, vnc, minimal, wayland, offscreen.
paź 12 19:27:31 Linux dbus-broker-launch[728]: Activation request for 'org.freedesktop.portal.Desktop' fai
led.
paź 12 19:27:31 Linux systemd-coredump[6268]: Failed to send coredump fd: Connection reset by peer
paź 12 19:27:31 Linux dhcpcd[600]: script_status: /usr/lib/dhcpcd/dhcpcd-run-hooks: Terminated
paź 12 19:27:32 Linux kernel: watchdog: watchdog0: watchdog did not stop!

Last edited by gerwazy (2024-10-12 20:04:07)

seth · 2024-10-12 19:57:16

Please use [code][/code] tags, not "quote" tags. Edit your post in this regard.

There's an IO error in the nvidia module, the relevant update was likely actually the kernel?
The reason is likely somewhat around https://bbs.archlinux.org/viewtopic.php … 5#p2087035 ie https://wiki.archlinux.org/title/NVIDIA … er_suspend ?

gerwazy · 2024-10-12 21:56:55

I found the cause – as I mentioned earlier, it’s related to the nvidia-utils update.

[2024-10-12T12:26:17+0200] [ALPM] upgraded nvidia-utils (560.35.03-5 -> 560.35.03-6)
[2024-10-12T12:26:17+0200] [ALPM-SCRIPTLET] If you run into trouble with CUDA not being available, run nvidia-modprobe first.
[2024-10-12T12:26:17+0200] [ALPM-SCRIPTLET] If you use GDM on Wayland, you might have to run systemctl enable --now nvidia-resume.service
[2024-10-12T12:26:17+0200] [ALPM-SCRIPTLET] If you use sleep on wayland, you might have to run systemctl enable nvidia-resume nvidia-hibernate nvidia-suspend

Steps I took:

- switch to kernel linux-lts and nvidia-lts
- nvidia-modprobe
- systemctl enable --now nvidia-resume.service
- systemctl enable nvidia-resume nvidia-hibernate nvidia-suspend

After these operations, everything returned to normal and is working properly. I might check tomorrow how the system behaves after switching back to the 6-11 kernel.

Last edited by gerwazy (2024-10-12 22:13:54)

cstdio · 2024-10-12 22:01:51

seth wrote:

One in particular looks interesting, try
amdgpu.aspm=0 amdgpu.runpm=0 amdgpu.bapm=0 pcie_aspm=off amdgpu.dpm=0
but dpm=0 in the past has lead to boot failures…

Just tried, can confirm dpm=0 leads to boot failure.

However I had another successful wake up. What happened is I booted into UEFI, switched to a fallback kernel image since dpm=0 prevented me from booting the main image, removed dpm=0 from kernel parameters of the main image, mkinitcpio and rebooted back to the main image. Right after this I was able to have a successful wake up, but the second try was unsuccessful.

I will try if switching kernel image may temporarily solve the issue later tonight and update it here.
Switching kernel image does not work
The only other thing I did before that successful wake up was added a password to UEFI settings. Maybe I should try changing something in the UEFI.

Last edited by cstdio (2024-10-12 22:52:40)

seth · 2024-10-13 07:58:07

gerwazy wrote:

I found the cause – as I mentioned earlier, it’s related to the nvidia-utils update.…
- systemctl enable --now nvidia-resume.service
- systemctl enable nvidia-resume nvidia-hibernate nvidia-suspend

Yes that perfectly fits my explanation in my previous post but not this thread as the OP doesn't have an nvidia GPU…

cstdio wrote:

The only other thing I did before that successful wake up was added a password to UEFI settings. Maybe I should try changing something in the UEFI.

cstdio previously wrote:

The only pattern I could notice is except the first lucky successful resume, all the other successful resume happened after I changed something to the PC (i.e. update motherboard BIOS, introduce a new keyboard). However I don't know if this is just coincidence.

Sp4rks: "Host: B650 GAMING X AX"
herbiejhopkins: "Product Name: B650E-I"
gerwazy: "Host: B550 GAMING X V2 (-CF) (but likely unrelated)"
cstdio: "DMI: ASRock B650M Pro RS WiFi/B650M Pro RS WiFi, BIOS 3.08 09/18/2024"

The only outlier would be Gasp0de with a thinkpad, watch_dog has an nvidia GPU and might fall into gerwazy's case.

gerwazy · 2024-10-13 10:30:26

I think it’s important to mention that Archlinux doesn’t support this motherboard by default after installation, and a manual fix needs to be applied. This might be of crucial importance here.

Archwiki

 systemctl status wakeup-disable_GPP0.service
○ wakeup-disable_GPP0.service - Fix suspend by disabling GPP0 sleepstate thingie
     Loaded: loaded (/etc/systemd/system/wakeup-disable_GPP0.service; enabled; preset: disabled)
     Active: inactive (dead) since Sun 2024-10-13 11:23:41 CEST; 52min ago
   Duration: 4ms
    Process: 586 ExecStart=/bin/bash -c echo GPP0 >> /proc/acpi/wakeup (code=exited, status=0/SUCCESS)

Last edited by gerwazy (2024-10-13 11:42:31)

herbiejhopkins · 2024-10-13 15:12:21

gerwazy wrote:

I think it’s important to mention that Archlinux doesn’t support this motherboard by default after installation, and a manual fix needs to be applied. This might be of crucial importance here.

Checked but I'm not disabling any wake up triggers on my machine. There seems to be reports of what looks to be the same issue on at Manjaro and Ubuntu.

Sp4rks · 2024-10-13 16:52:10

gerwazy wrote:

I think it’s important to mention that Archlinux doesn’t support this motherboard by default after installation, and a manual fix needs to be applied. This might be of crucial importance here.

I've not applied that fix before, and sleep/resume was working fine before 6.11.

So I switched back to baseline kernel (from lts) to try this, see if it had an effect.

First sleep/resume was OK. So I was hopeful it had worked, but second and third time, the issue came back.

seth · 2024-10-13 17:55:37

That fix will prevent some (nvme) interrupt from spuriously waking the system - it will not do much about the system not waking up properly.
gerwazy's situation is down to a misconfiguration of the nviida VRAM preservation and has been a problem before, but might have gotten worse because of the various framebuffer related issues in 6.11 (that may very well in a different or similar way be the ultimate cause of this board/amdgpu specific issues)

Most others in this thread have the same/similar board and according to cstdio, every time he touches the UEFI, the next sleep cycle works…

cstdio · 2024-10-13 22:58:33

seth wrote:

... according to cstdio, every time he touches the UEFI, the next sleep cycle works…

small correction, I was speculating that because almost every successful sleep cycle was after I touched UEFI. However I just explicitly tried several tests (changing UEFI password, change memory settings) and they didn't work now...

I am confused.

Last edited by cstdio (2024-10-13 23:00:43)

pinguinwizard · 2024-10-14 00:17:55

Can confirm the problem happening.

I have:
- Lenovo Thinkpad P15v Gen3
- AMD Ryzen CPU
- Integrated GPU AMD Radeon 680M
- NVME disk
- I use Plasma v6 with Wayland (with SDDM)
- standard installation: Grub, no extra kernel parameters (all default), all packages up-to-date (kernel 6.11.3.arch1-1) today (2024-10-13)
- only thing that is probably not very common is that I use system-homed (for encrypting home partition which auto mounts after login), but no encryption on root partition

After "wakeup" the screen is sometimes black, sometimes with clock, sometimes with (not moving) cursor, not reacting to any key, had to hard reboot (with holding power button for 5 secs).
No messsage in journalctl AFTER wakeup. (no error or suspicious message in journal during suspend)

After downgrading kernel (packages linux and linux-headers to 6.10.10-arch1-1 - these were latest 6.10 in my local cache) everything works.

Last edited by pinguinwizard (2024-10-14 00:21:18)

herbiejhopkins · 2024-10-14 08:02:08

I read on kernel bugzilla that this could be bluetooth/wifi related. I don't use wifi on my affected machine (it's always disabled) but I do use bluetooth. If I disable bluetooth my machine can then suspend/resume successfully!

Created a quick systemd service to automate this and suspend/resume seems to be working well again.

[Unit]
Description=Disable wifi/bluetooth on suspend
Before=suspend.target hibernate.target hybrid-sleep.target suspend-then-hibernate.target

[Service]
ExecStart=/usr/bin/rfkill block all

[Install]
WantedBy=suspend.target hibernate.target hybrid-sleep.target suspend-then-hibernate.target

Last edited by herbiejhopkins (2024-10-14 10:25:59)

cstdio · 2024-10-14 17:13:29

Unfortunately for me blocking wifi and bluetooth only makes the next sleep cycle work. Even if I rfkill again after waking up the second sleep cycle is still stuck and needs a reboot.

seth · 2024-10-14 21:43:23

mt7921e also shows up in https://bbs.archlinux.org/viewtopic.php?id=300126
Instead of rfkill, can you disable the NIC in the firmware (BIOS/UEFI)? resp. try to blacklist mt7921e ?

If the chip ends up jamming the bus, that'd explain why nothing else (GPU, nvme) works anymore…

Gasp0de · 2024-10-14 21:43:39

Sorry for not responding over the weekend, my system runs from an nvme drive as well.

Sp4rks · 2024-10-14 22:23:40

herbiejhopkins wrote:

I read on kernel bugzilla that this could be bluetooth/wifi related. I don't use wifi on my affected machine (it's always disabled) but I do use bluetooth. If I disable bluetooth my machine can then suspend/resume successfully!

Wow, that worked for me. Good find.

Just wanted to document my steps for others. Mobo B650 GAMING X AX has bluetooth and wifi, but I only use the bluetooth on mine. So, initially did a couple sleep/resume cycles with just:

systemctl bluetooth disable

And that worked ok. Then I copied and enabled herbiejhopkins's rfkill service, tested that too, works perfectly.

Arch Linux

#26 2024-10-09 20:58:10

Re: Black screen after resuming from sleep

#27 2024-10-09 21:11:13

Re: Black screen after resuming from sleep

#28 2024-10-09 21:25:20

Re: Black screen after resuming from sleep

#29 2024-10-10 17:28:36

Re: Black screen after resuming from sleep

#30 2024-10-11 12:54:35

Re: Black screen after resuming from sleep

#31 2024-10-12 08:47:49

Re: Black screen after resuming from sleep

#32 2024-10-12 15:19:48

Re: Black screen after resuming from sleep

#33 2024-10-12 17:08:08

Re: Black screen after resuming from sleep

#34 2024-10-12 18:43:35

Re: Black screen after resuming from sleep

#35 2024-10-12 19:33:34

Re: Black screen after resuming from sleep

#36 2024-10-12 19:57:16

Re: Black screen after resuming from sleep

#37 2024-10-12 21:56:55

Re: Black screen after resuming from sleep

#38 2024-10-12 22:01:51

Re: Black screen after resuming from sleep

#39 2024-10-13 07:58:07

Re: Black screen after resuming from sleep

#40 2024-10-13 10:30:26

Re: Black screen after resuming from sleep

#41 2024-10-13 15:12:21

Re: Black screen after resuming from sleep

#42 2024-10-13 16:52:10

Re: Black screen after resuming from sleep

#43 2024-10-13 17:55:37

Re: Black screen after resuming from sleep

#44 2024-10-13 22:58:33

Re: Black screen after resuming from sleep

#45 2024-10-14 00:17:55

Re: Black screen after resuming from sleep

#46 2024-10-14 08:02:08

Re: Black screen after resuming from sleep

#47 2024-10-14 17:13:29

Re: Black screen after resuming from sleep

#48 2024-10-14 21:43:23

Re: Black screen after resuming from sleep

#49 2024-10-14 21:43:39

Re: Black screen after resuming from sleep

#50 2024-10-14 22:23:40

Re: Black screen after resuming from sleep

Board footer