You are not logged in.
Hello everyone,
I'm having a serious issue with my Lenovo Legion 5 (Arch Linux) since kernel 6.11. Whenever my laptop goes into suspend (whether triggered by the system or by closing the lid), it won't properly resume. When I try to wake it up, the screen turns on, keyboard lights up, but only shows a blinking dash cursor on a black screen.At that point, the system is completely frozen. No keypresses work, including switching TTYs or Ctrl+Alt+Del, and I have to do a hard power-off.
Until recently, I was using the LTS kernel (which was on a 6.7 or earlier version) as a workaround, and that worked fine. However, after the LTS kernel updated to 6.12, I'm now seeing the same issue there too. This has become really disruptive, because I'm constantly forced to do a hard shutdown if the laptop suspends.
My Setup:
- Lenovo Legion 5
- Arch Linux (fully up to date)
- Kernel 6.13.1-arch1-1 (Same issue started on 6.11+)
- CPU: 12th Gen Intel(R) Core(TM) i7-12700H
- GPUs: (1) Intel Alder Lake-P GT2 Iris Xe Graphics; (2) NVIDIA GA104M GeForce RTX 3070 Mobile / Max-Q
- Nvidia driver: nvidia 570.86.16-2. Also tried with nvidia-dkms under linux-lts-6.12 - same result
- Hyprland on Wayland. Also tried 'systemctl suspend' under vanilla TTY - same result
Configs and logs
- [cat /proc/cmdline](https://pastes.io/cmdline-45)
- [journalctl -b -1](https://pastes.io/journalctl-b-1)
- [journalctl -b -1 -k](https://pastes.io/journalctl-b-1-k)
- [dmesg](https://pastes.io/dmesg-9306)
- [/etc/mkinitcpio.conf](https://pastes.io/mkinitcpioconf)
- [/etc/modprobe.d/nvidia.conf](https://pastes.io/nvidiaconf)
- `nvidia-persistanced.service` disabled (tried enabling too - same result)
- [inxi -Fxxxz output](https://pastes.io/inxi-14351)
Troubleshooting Steps Tried:
- Using various systemd sleep modes (deep, s2idle).
- Downgrading to older kernels (6.10 or older) fixes the issue, but those versions are no longer readily available.
- Tested with stock arch kernel vs. LTS kernel, same results on 6.11+.
No luck so far. The freeze on resume is consistent regardless of how I trigger suspend.
Has anyone else run into this issue or found a workaround/fix? Any help is greatly appreciated!
Thank you in advance.
Last edited by seethesky (2025-02-05 09:53:19)
Offline
Sorry, I don't have any advice, but thought I'd raise it that I have the same issue.
I've followed roughly the same steps as you so far (downgraded the kernel to 6.6.72 LTS) to be able to suspend.
I'm running this on a Lenovo x1 nano G2, Core i7-1260p, Samsung 980 NVMe.
Offline
Thanks for the support. I am puzzled by the lack of responses and how this issue slipped through into 6.12 LTS. I have downgraded to 6.7 LTS, but this cannot go on forever.
Offline
You can setup sysrq and do REISUB key combination slowly when that happens.
Also you may try to lower sleep state in BIOS (if you have related settings) or/and disable suspend.
Unfortunately, I don't see any clue in logs - it probably don't save anything after sleep.
Offline
I also have experienced this, I have additionally tried the following, all without a positive result:
- Disabling Wifi and Bluetooth before suspending
- NVreg_PreserveVideoMemoryAllocations=1
- setting acpi_os_name="Linux"
- setting acpi_os_name="Windows"
- setting "intel_idle.max_cstate=1"
I think I've tried some other possible solutions too, but there's been so many at this point I have lost track from all the different methods aside from just what browser tabs I have open.
My current solution is just to use an older Linux version, but it is very important that a proper detailed bug report is made for this, if that has not been done already, as this is practically breaking userspace.
Offline
I also have experienced this, I have additionally tried the following, all without a positive result:
My current solution is just to use an older Linux version, but it is very important that a proper detailed bug report is made for this, if that has not been done already, as this is practically breaking userspace.
I have also downgraded to 6.7 LTS because of this. The problem with making a proper bug report is that I can't pin it down in the logs and can't set out how to reproduce the issue.
Offline
I've had some time to look at this again today, and have narrowed it down to the deep sleep, but haven't got a root cause. I've mainly used the kernel.org info on sleep.
I thought I'd solved it (and wrote it up here) but it looks as though I'd made an error in process and booted the 'stable' kernel.
For now, I'm working around it using systemd sleep config to use s2idle rather than deep:
# /etc/systemd/sleep.conf
[Sleep]
MemorySleepMode=s2idleNot perfect but it allows me to use the current kernel and still sleep.
I plan on looking at firmware next time I get chance. Also, on kernel.org, there's some info round using the RTC to capture issues, under the pm_state heading, so it's worth reading that to see if we can get some useful info from that too.
Last edited by technononsense (2025-02-12 16:18:22)
Offline
For now, I'm working around it using systemd sleep config to use s2idle rather than deep:
s2idle does not work for me. Tried just now with the latest linux kernel. Same result on waking up - blinking dash cursor and the freeze. I'm going to try with fresh install when I have time.
Offline
I am now convinced this is truly bizarre and hard-coded into the kernel. I created a clean Live USB and tested suspend before installing extra packages, nvidia drivers and without editing many configs. Then tested after nvidia and with minimal config. The result is the same - screen wakes up, drives and fans wake up, keyboard lights up, but everything is completely unresponsive. I would have thought this was hardware problem, but this does not happen on 6.6.72-1-lts.
Offline
Hi, I've also been having the same issue for the past months, now I am on linux-6.12.12.
Is there an issue on the linux kernel bugtracker already?
Offline
I have had this same exact problem, for some reason the problem dissapears obviously when using LTS but also weirdly when I have a swap partition( as the problem only occured only when I did not have a swap) or when I use an automated script like archinstall, these are my solutions and I don't know why they work
Last edited by Omarz2012 (2025-02-26 19:22:11)
Offline
Having lived with s2idle for months, but unable to suspend, I spent some time on this again last week & the weekend. I've managed to isolate an intel module as the cause of my issues. In my case its the i2c_ljca module. Tested using:
cat /sys/power/mem_sleep
s2idle [deep]
modprobe -r i2c_ljca && systemctl sleepBoth sleep & hibernate now work as expected. I'll add it to systemd pre|post sleep|hibernate hooks now.
I had wondered whether it would be intel related, due to how poor the driver situation is for camera etc.
Last edited by technononsense (2025-07-07 11:49:13)
Offline