1. Please describe the problem: The systems fail to resume from sleep. I have two systems, fresh installs on both, both KDE plasma. 1) ASRock x570 AMD 5950X + NVidia 4090 FE - Bazzite 20260101 - Linux 6.17.7-ba22.fc43.x86_64 2) Minisforum MS-S1 (AMD AI MAX 395+), integrated GPU. Fedora 43 - Linux fedora 6.17.12-300.fc43.x86_64 On BOTH machines, it's possible to SSH into them, however a `sudo shutdown -r now` does NOT reboot the machine, you always have to force-reset them. Both have working networking/wifi/sound/usb/keyboard (caps lock works). Both systems have been tried without any peripherals other than keyboard + mouse + monitor. On machine 1, sometimes it will sleep and resume ONCE, but the second time it will always fail to resume, no display output, all monitors remain on standby. On machine 2, the machine will sleep but will NEVER resume. again the monitor remains on standby. 2. What is the Version-Release number of the kernel: See above. 3. Did it work previously in Fedora? If so, what kernel version did the issue *first* appear? Old kernels are available for download at https://koji.fedoraproject.org/koji/packageinfo?packageID=8 : Not tried older versions of Fedora, these are new installs. 4. Can you reproduce this issue? If so, please provide the steps to reproduce the issue below: Yes, always. To replicate the problem simply start the machine and then sleep it, wake it using the keyboard, if it wakes correctly, sleep it again and try waking again. 5. Does this problem occur with the latest Rawhide kernel? To install the Rawhide kernel, run ``sudo dnf install fedora-repos-rawhide`` followed by ``sudo dnf update --enablerepo=rawhide kernel``: Unknown. 6. Are you running any modules that not shipped with directly Fedora's kernel?: No, changes made to the kernel or driver configuration, everything standard. 7. Please attach the kernel logs. You can get the complete kernel log for a boot with ``journalctl --no-hostname -k > dmesg.txt``. If the issue occurred on a previous boot, use the journalctl ``-b`` flag. Have logs, will try and attach. Form has no attach button... Additionally, there are many other users reporting similar things, just look at recent bug reports with the word `resume` in them. Reproducible: Always
Created attachment 2121003 [details] full journalctl log from machine 1 (ASRock x570)
Created attachment 2121004 [details] full journalctl log from machine 2 (Minisforum MS-S01)
Additionally, both systems report this: machine 1: ``` Jan 03 22:01:15 kwin_wayland[2726]: Pageflip timed out! This is a bug in the nvidia-drm kernel driver Jan 03 22:01:15 kwin_wayland[2726]: Please report this at https://forums.developer.nvidia.com/c/gpu-graphics/linux ``` machine 2: ``` Jan 03 22:15:23 kwin_wayland[2506]: Pageflip timed out! This is a bug in the amdgpu kernel driver Jan 03 22:15:23 kwin_wayland[2506]: Please report this at https://gitlab.freedesktop.org/drm/amd/-/issues ``` I have also tried various workarounds on both machines, such as enabling/disable global c-states in the bios, using kernel args like `mem_sleep_default=deep` and `pcie_aspm=off` independantly, but to no avail.
Today, for fun, I updated Machine 2 (The Minisforum MS-S1, AMD Stryx Halo, AMD AI MAX 395+ CPU, using simply `sudo dnf update` This upgraded the kernel from: ``` Linux fedora 6.17.12-300.fc43.x86_64 ``` to ``` Linux fedora 6.18.4-200.fc43.x86_64 #1 SMP PREEMPT_DYNAMIC Thu Jan 8 17:35:49 UTC 2026 x86_64 GNU/Linux ``` ``` $ cat /proc/version Linux version 6.18.4-200.fc43.x86_64 (mockbuild@bc1e0bf216f84bd4841c551523010be9) (gcc (GCC) 15.2.1 20251211 (Red Hat 15.2.1-5), GNU ld version 2.45.1-1.fc43) #1 SMP PREEMPT_DYNAMIC Thu Jan 8 17:35:49 UTC 2026 ``` Many other packages were also updated. On this machine, s2idle is used. ``` $ cat /sys/power/mem_sleep [s2idle] ``` The system will now go to sleep and will also resume from sleep multiple times in a row. I will try and update Machine 1 to the 6.18.4-200 kernel, which is more complicated since it uses Bazzite, and report back. I will also attach a log from Machine 1 which includes the 2 sleep resume cycles.
Created attachment 2121908 [details] machine-2-20260112-two-sleeps.txt full journalctl log from machine 2 after upgrading to kernel 6.18.4-200.
At the time the kernel update and successful sleep/resume cycles, Minisforum MS-S1 BIOS, version 1.05, had 'Global C-States: DISABLED'. I reset that back to: Global C-States: Auto' and performed the test again, however the machine would not go to sleep, it just woke up immediately, attempting to sleep again resulted in the display locking and unresponsive keyboard (no caps lock activity) and no mouse, the system was available over the network so I grabbed another log. While I was typing this the system became responsive again and I was able to login. The system still does not sleep though.
Created attachment 2121909 [details] machine-2-20260112-02-global-cstates-auto.txt log from machine 2, with Global C-States set to AUTO in the bios. Machine does not sleep at all.
Related problem: When I close the screen of my laptop onto the keyboard, it goes into sleep mode, but I can't always restart it: the screen remains black and does not display the password prompt. However, if I lock the screen (Super+L) or wait a long time for the computer to go into sleep mode, I can always get the password prompt. Hardware: MacBook Air OS: Fedora 43 / GNOME
Related problem: When I lock the screen (Super+L) or wait a long time for the computer to go into sleep mode, one of my screen remains black. When I lock again and unlock back normal. Hardware: AMD Ryzen 5 PRO 2600 x 12 / AMD Radeon RX 550 Series OS: Fedora 43 / GNOME