[SOLVED] Immediate Wake After Suspend

sylv256 · 2025-02-18 03:52:53

Hello, I've been having trouble suspending only in Linux; in Windows it works just fine. I suspect it is my motherboard or the graphics drivers.

Here's my hyfetch:

                                         sylv@protogen 
                   -`                    ------------- 
                  .o+`                   OS: Arch Linux x86_64 
                 `ooo/                   Host: B650 GAMING X AX V2 
                `+oooo:                  Kernel: 6.13.2-arch1-1 
               `+oooooo:                 Uptime: 1 hour, 34 mins 
               -+oooooo+:                Packages: 838 (pacman) 
             `/:-:++oooo+:               Shell: fish 3.7.1 
            `/++++/+++++++:              Resolution: 1920x1080 
           `/++++++++++++++:             DE: GNOME 47.4 (wayland) 
          `/+++ooooooooooooo/`           WM: Mutter 
         ./ooosssso++osssssso+`          WM Theme: Adwaita 
        .oossssso-````/ossssss+`         Theme: Adwaita [GTK2/3] 
       -osssssso.      :ssssssso.        Icons: Adwaita [GTK2/3] 
      :osssssss/        osssso+++.       Cursor: Bibata Modern v1.0.2 X11 [GTK2/3] 
     /ossssssss/        +ssssooo/-       Terminal: kitty 
   `/ossssso+/:-        -:/+osssso+-     Terminal Font: monospace 11.0 
  `+sso+:-`                 `.-/+oso:    CPU: AMD Ryzen 7 7700X (16) @ 5.5GHz 
 `++:.                           `-/+/   GPU: AMD ATI Raphael 
 .`                                 `/   GPU: NVIDIA GeForce RTX 3060 Lite Hash Rate 
                                         Memory: 7.50 GiB / 30.46 GiB (24%) 
                                         Network: Wifi 
                                         BIOS: American Megatrends International, LLC. 5.35 (02/07/2025)

This is my journalctl after attemping to suspend multiple times using systemctl suspend. I also tried echo -n "mem" | sudo tee /sys/power/state, but all that seemed to do was briefly shut off my keyboard.
https://0x0.st/8q3R.txt

Things I've tried:
Disabling everything in /proc/acpi/wakeup.
Using acpi_osi=! acpi_osi="Windows 2022" (this caused the fans to stop spinning and a black screen).
Disabling every "Wake On X" feature in my BIOS. There also is no Wake On LAN option in my BIOS settings or I could not find it.
Disabling hibernation on Windows.
This helped with the weird suspend hangs.

Similar Issues:
https://bbs.archlinux.org/viewtopic.php?pid=2004037
https://bbs.archlinux.org/viewtopic.php?id=292633
https://bbs.archlinux.org/viewtopic.php?id=302440

Edit:
I also upgraded the BIOS to the latest version (F33a) and that did nothing new except make some parts of the BIOS UI display in Spanish.

Solution

I updated my BIOS to version F33a and checked the DSDT for strings related to Windows. I then saw Windows 2015 was the latest supported _OSI string and added the following to my kernel parameters:

acpi_osi="!Windows 2015"

Finally, the computer suspended completely, albeit with Wake-On-X functionality enabled. Enabling ErP mode disabled Wake-On-X. I hope this helps anyone else using a Gigabyte B650 GAMING X AX V2 motherboard or similar.

If this fixed your issue, please add your motherboard to the section on the wiki.

Last edited by sylv256 (2025-02-20 04:57:13)

seth · 2025-02-18 09:05:36

Feb 17 21:13:11 protogen kernel: ACPI: PM: Preparing to enter system sleep state S3
Feb 17 21:13:11 protogen kernel: ACPI: PM: Saving platform NVS memory
Feb 17 21:13:11 protogen kernel: Disabling non-boot CPUs ...
Feb 17 21:13:11 protogen kernel: smpboot: CPU 15 is now offline
…
Feb 17 21:13:11 protogen kernel: smpboot: CPU 8 is now offline
Feb 17 21:13:11 protogen kernel: Spectre V2 : Update user space SMT mitigation: STIBP off
Feb 17 21:13:11 protogen kernel: smpboot: CPU 7 is now offline
…
Feb 17 21:13:11 protogen kernel: smpboot: CPU 1 is now offline
Feb 17 21:13:11 protogen kernel: ACPI: PM: Low-level resume complete
…
Feb 17 21:13:11 protogen kernel: Spurious APIC interrupt (vector 0xFF) on CPU#9, should never happen.

Does s2idle fare any better than S3?

Using acpi_osi=! acpi_osi="Windows 2022" (this caused the fans to stop spinning and a black screen).

Feb 17 21:11:08 protogen kernel: Command line: root=UUID=63b4e46d-d336-4401-8a2d-a0537f32c3a7 rw acpi_osi="Windows 2022"

So the journal simply covers that condition? How does the system behave w/o faking any windows version?

Disabling hibernation on Windows.

Incl. the implicit fast-start etc, 3rd link below and rebooted windows and linux twice for voodo reasons?

sylv256 · 2025-02-18 11:27:19

seth wrote:

Does s2idle fare any better than S3?

It does sleep until the power button is pressed like normal, but it's not very useful for a desktop with a bunch of lights and loud fans (I need better cooling eventually).

Seth wrote:

Using acpi_osi=! acpi_osi="Windows 2022" (this caused the fans to stop spinning and a black screen).
Feb 17 21:11:08 protogen kernel: Command line: root=UUID=63b4e46d-d336-4401-8a2d-a0537f32c3a7 rw acpi_osi="Windows 2022"
So the journal simply covers that condition? How does the system behave w/o faking any windows version?

I'm not sure what you mean, but I manually entered that and it had no difference than the times I had tested without it. I already tried acpi_osi=! with acpi_osi="Windows 2022" which caused my computer's fans to stop spinning and a black screen.

seth wrote:

Disabling hibernation on Windows.
Incl. the implicit fast-start etc, 3rd link below and rebooted windows and linux twice for voodo reasons?

I only used powercfg /H off and then shut down windows, but I didn't shut down Windows twice and then Linux twice. I'll try that, and if that doesn't work, I'll try messing with my BIOS settings again.

seth · 2025-02-18 12:47:52

not very useful for a desktop with a bunch of lights and loud fans

s2idle is supposed to shut all of that down (some status LED aside, maybe) - if that doesn't happen it basically doesn't work (either).

I meant whether the journal is actually for a boot that "caused the fans to stop spinning and a black screen".
Or is "caused the fans to stop spinning and a black screen" the default state (ie. the system doesn't properly boot at all) with those parameters?

Since more users of the same board seem affected and you've already disabled all acpi/wakeup devices, I've limited hope, but do you use a wireless keyboard/mouse?

sylv256 · 2025-02-18 17:09:56

seth wrote:

I meant whether the journal is actually for a boot that "caused the fans to stop spinning and a black screen".
Or is "caused the fans to stop spinning and a black screen" the default state (ie. the system doesn't properly boot at all) with those parameters?

This journal is the one that boots. The addition of only acpi_osi="Windows 2022" did nothing in this case. I don't actually have a journal for the acpi_osi=! acpi_osi="Windows 2022" run because it seems to never boot. I have seen examples on the wiki like acpi_os="!Windows <YEAR>" which supposedly function the same; maybe the syntax changed? I'll try that too when I get home today. I've also found something rather concerning that may be related since Gigabyte et al. are known for breaking things on Linux.

seth wrote:

Since more users of the same board seem affected and you've already disabled all acpi/wakeup devices, I've limited hope, but do you use a wireless keyboard/mouse?

I have a wired USB keyboard and a wireless USB mouse.

Last edited by sylv256 (2025-02-18 17:58:47)

seth · 2025-02-18 20:50:18

The implication would be to try w/o the wireless mouse, but again: limted hope given the circumstances.

You could try adding "noapic" to the kernel parameters, but that's hardly a soltion even if it works.
Does "iommu=soft" help?

sylv256 · 2025-02-18 22:01:41

I tried booting with some new ACPI parameters based on suggestions on the Wiki and in my DSDT and it ended up black-screening me. I had to reset the PC to get it working again because the monitor somehow froze. Unfortunately, no luck.

http://0x0.st/8bvw.txt

Also, the iommu=soft parameter did nothing. ErP as well.

Last edited by sylv256 (2025-02-18 22:31:37)

seth · 2025-02-19 16:30:46

The boot in that journal neither use noapic nor iommu=soft?

acpi_os_name="Microsoft Windows NT" acpi_osi="Windows 2015"

This doesn't work.

acpi_osi=! acpi_osi="Windows 2015"

https://learn.microsoft.com/en-us/windo … inacpi-osi

The "=!" is necessary to clear the list and "Windows NT" is not a proper entry - I don't think ACPI was even concepted when NT was a relevant OS.

Last edited by seth (2025-02-19 16:30:56)

sylv256 · 2025-02-19 21:17:57

seth wrote:

The boot in that journal neither use noapic nor iommu=soft?
acpi_os_name="Microsoft Windows NT" acpi_osi="Windows 2015"
This doesn't work.
acpi_osi=! acpi_osi="Windows 2015"
https://learn.microsoft.com/en-us/windo … inacpi-osi
The "=!" is necessary to clear the list and "Windows NT" is not a proper entry - I don't think ACPI was even concepted when NT was a relevant OS.

I got it working.. somehow. Here's the dmesg and journalctl logs. I'm going to do more testing to see what exactly worked.

Here's the kernel parameters:

root=UUID=63b4e46d-d336-4401-8a2d-a0537f32c3a7 rw acpi_osi="!Windows 2015" acpi_os_name="Microsoft Windows"

Last edited by sylv256 (2025-02-19 21:18:57)

sylv256 · 2025-02-19 21:25:42

I can confirm that the issue is now solved. I'll add the solution to the top of the post.

congard · 2025-02-19 22:50:03

Thank you very much, your fix works also for bios version F31.

shouya · 2025-02-20 02:28:30

After having to hibernate all these days I'm so happy to finally enjoy a good night's sleep. Thank you sylv256!

Arch Linux

#1 2025-02-18 03:52:53

[SOLVED] Immediate Wake After Suspend

Solution

#2 2025-02-18 09:05:36

Re: [SOLVED] Immediate Wake After Suspend

#3 2025-02-18 11:27:19

Re: [SOLVED] Immediate Wake After Suspend

#4 2025-02-18 12:47:52

Re: [SOLVED] Immediate Wake After Suspend

#5 2025-02-18 17:09:56

Re: [SOLVED] Immediate Wake After Suspend

#6 2025-02-18 20:50:18

Re: [SOLVED] Immediate Wake After Suspend

#7 2025-02-18 22:01:41

Re: [SOLVED] Immediate Wake After Suspend

#8 2025-02-19 16:30:46

Re: [SOLVED] Immediate Wake After Suspend

#9 2025-02-19 21:17:57

Re: [SOLVED] Immediate Wake After Suspend

#10 2025-02-19 21:25:42

Re: [SOLVED] Immediate Wake After Suspend

#11 2025-02-19 22:50:03

Re: [SOLVED] Immediate Wake After Suspend

#12 2025-02-20 02:28:30

Re: [SOLVED] Immediate Wake After Suspend

Board footer