You are not logged in.
This issue has started occurring since I recently swapped my psu and gpu. What happens is the PC shuts down everything stops it even disconnects my mouse and keyboard. But the power led, gpu, and fans, and gpu output to monitors stays on. It only seems to occur while the PC has been running for 3+-. If I shutdown soon after booting up the PC it shuts down like it is supposed to. For now I am shutting it down by holding the power button. HDD led stops blinking so I just hope they are off . The command I use for shutdown is systemctl poweroff.
PC Specs:
CPU: Intel I5-10400f
MOBO: GIGABYTE B460M DS3H
GPU: Intel Arc B580 Limited Edition
RAM: HyperX 32 GB KIT DDR4 3200 MHz CL16
Sata Storage: 8TB hdd,4TB hdd,240gb ssd,2TB ssd
NVME Storage: 1TB ssd
PSU: Corsair RM1000e
Before the upgrade I was using an Nvidia GTX 1060 6gb and a EVGA 500w psu.
I also use a Bluetooth adapter in the back of my pc.
The sata drives are used for linux. The nvme one is for windows. I use Grub to dualboot.
As for things I tried to fix this is:
disabling watchdog
updating Bios to the latest one
disabling secure boot
disabling windows fast boot
Using Alt + Sysrq R E I S U B (doesn't work)
Here is my paste bin for:
Any help is really appreciated.
Last edited by ikorus (2025-04-14 20:23:10)
Offline
Bunch of crashes in the xe driver, I'd say test the 6.13 kernel by grabbing it from testing, or wait a bit for it to make it into the stable repos
Offline
Bunch of crashes in the xe driver, I'd say test the 6.13 kernel by grabbing it from testing, or wait a bit for it to make it into the stable repos
Ok. I installed 6.13 from core-testing. Will check for errors tomorrow. I'll post a new journal if it occurs again. If it doesn't within a week I'll mark as solved.
Offline
Alright so today it happened again. But this time it actually turns off the monitors and terminates the monitor output signal. But the PC case power LED and the fans stay on.
I also see a lot of these:
Feb 03 12:00:41 IkorusPC xdg-desktop-por[1209]: Backend call failed: Cannot invoke method; proxy is for the well-known name org.freedesktop.ScreenSaver without an owner, and proxy was constructed with the G_DBUS_PROXY_FLAGS_DO_NOT_AUTO_START flag
Should I worry about these ?
I saw in the journal that SDDM was throwing a few errors while shutting down so I disabled it.
Feb 02 23:28:25 IkorusPC sddm[812]: Authentication error: SDDM::Auth::ERROR_INTERNAL "Process crashed"
Feb 02 23:28:25 IkorusPC sddm[812]: Authentication error: SDDM::Auth::ERROR_INTERNAL "Process crashed"
Here is the journal from this last fail:
Again thank you in advance.
Offline
I have nearly the same CPU: i5-10400 (non-f), and nearly the same mobo: Gigabyte B460M D3H (non-S) and am encountering the same issues. I don't use SDDM, so my guess is that the issues are not related to that. One additional thing I tried is switching to the LTS kernel, which did not help. My next guess is the zen kernel...
Other than the mobo and cpu we don't seem to have much in common, so maybe the issue is related to this?
Offline
To be honest for me, it didn't happen in almost a week, but I haven't been using my PC as much lately.
If you can post
sudo journalctl -b -1
Change -1 to -n, of which n is the last boot sequence in which the freeze happened.
For example, let's say the freeze happened 3 boots back, so I would do
sudo journalctl -b -3
So that we can at least compare errors in journals.
For me, the freeze hasn't happened lately. Although this could just be random as I haven't been using my PC much for the past week.
I will give it one more week. If the freeze doesn't happen, I'll mark it as solved.
Offline
As luck would have it yesterday, it got stuck again. As mentioned previously, it seems to happen randomly. Also, I noticed that pastebin removed my second journal. So I searched and mods here recommend uploading to 0x0.st I'll just do that from now on.
Latest journal:
I can see a few errors/failures during the shutdown process.
Feb 15 23:11:06 IkorusPC systemd[870]: xdg-desktop-portal-gtk.service: Main process exited, code=exited, status=1/FAILURE
Feb 15 23:11:06 IkorusPC systemd[870]: xdg-desktop-portal-gtk.service: Failed with result 'exit-code'
Feb 15 23:11:06 IkorusPC systemd[870]: hyprpolkitagent.service: Main process exited, code=exited, status=255/EXCEPTION
Feb 15 23:11:06 IkorusPC systemd[870]: hyprpolkitagent.service: Failed with result 'exit-code'.
Feb 15 23:11:06 IkorusPC systemd[870]: xdg-desktop-portal-hyprland.service: Main process exited, code=dumped, status=11/SEGV
Feb 15 23:11:06 IkorusPC systemd[870]: xdg-desktop-portal-hyprland.service: Failed with result 'core-dump'.
Feb 15 23:11:06 IkorusPC dbus-broker-launch[654]: Activation request for 'org.freedesktop.nm_dispatcher' failed.
Feb 15 23:11:06 IkorusPC dbus-broker-launch[654]: Activation request for 'org.bluez' failed.
Feb 15 23:11:06 IkorusPC dbus-broker-launch[654]: Activation request for 'org.bluez' failed.
Feb 15 23:11:06 IkorusPC pipewire[1296]: spa.alsa: 'hdmi:1': playback open failed: No such file or directory
Feb 15 23:11:06 IkorusPC wireplumber[1297]: s-monitors: Failed to create alsa_output.pci-0000_04_00.0.hdmi-stereo: Object activation aborted: PipeWire proxy destroyed
Feb 15 23:11:07 IkorusPC dbus-broker-launch[654]: Activation request for 'org.freedesktop.nm_dispatcher' failed.
I compared the 3 journals and these errors/fails occur in all journals.
Feb 01 22:56:48 IkorusPC systemd[1015]: hyprpolkitagent.service: Main process exited, code=exited, status=255/EXCEPTION
Feb 01 22:56:48 IkorusPC systemd[1015]: hyprpolkitagent.service: Failed with result 'exit-code'.
Feb 01 22:56:48 IkorusPC systemd[1015]: xdg-desktop-portal-gtk.service: Main process exited, code=exited, status=1/FAILURE
Feb 01 22:56:48 IkorusPC systemd[1015]: xdg-desktop-portal-gtk.service: Failed with result 'exit-code'.
Feb 01 22:56:50 IkorusPC dbus-broker-launch[640]: Activation request for 'org.freedesktop.nm_dispatcher' failed.
Just for comparison, I'll post a journal below where the PC shuts off normally:
Normal shutdown journal
Offline
Here is a failed one of mine: journalctl. Not a very clean one with a bunch of core dumps from Steam and Remmina (RDP client).
With respect to your errors:
I dont have hyprpolkitagent installed
I've masked xdg-desktop-portal-gtk.service with:
systemctl --user mask xdg-desktop-portal-gtk.service
org.freedesktop.nm_dispatcher seems related to flatpak which I also did not have installed
Offline
Based on this line in both our logs:
ACPI FADT declares the system doesn't support PCIe ASPM, so disable it
I've added pcie_aspm=off to my GRUB config as kernel parameter:
GRUB_CMDLINE_LINUX_DEFAULT="loglevel=3 dis_ucode_ldr nowatchdog sysrq_always_enabled=1 pcie_aspm=off"
With the LTS kernel I've had a few succesful shutdowns in a row since. In the Gigabyte UEFI settings, Platform Power Management is set to the default value of disabled (this is what Gigabyte seems to call ASPM).
[user1@arch ~]$ uname -a
Linux arch 6.12.13-1-lts #1 SMP PREEMPT_DYNAMIC Sat, 08 Feb 2025 15:16:04 +0000 x86_64 GNU/Linux
Offline
Based on this line in both our logs:
ACPI FADT declares the system doesn't support PCIe ASPM, so disable it
I've added pcie_aspm=off to my GRUB config as kernel parameter:
GRUB_CMDLINE_LINUX_DEFAULT="loglevel=3 dis_ucode_ldr nowatchdog sysrq_always_enabled=1 pcie_aspm=off"
I tried that since yesterday. Seem to work so far, but do to the random nature of these freezes I'll give it a week again. I'm gonna be actively monitoring my journal for errors and patterns that may cause the shutdown freezing.
In the meantime I fixed the:
Feb 03 12:00:41 IkorusPC xdg-desktop-por[1209]: Backend call failed: Cannot invoke method; proxy is for the well-known name org.freedesktop.ScreenSaver without an owner, and proxy was constructed with the G_DBUS_PROXY_FLAGS_DO_NOT_AUTO_START flag
By using Hypridle. So at least that's that.
Offline
So no luck with pcie_aspm=off, the same issue remains. I've now switched to pcie_aspm=force after enabling Platform Power Management, PEG ASPM, PCH ASPM and DMI ASPM in the BIOS. So far so good... Still the same issue.
Last edited by R290 (2025-02-25 19:41:14)
Offline
BIG DISCLAIMER HERE:
Do not do any of the things I did. I am a certified electrician. And I know what I can and can't do when it comes to electricity.
Right I'm back. No new info. PC still does the same stuff. But I thought that it might be a good idea to list the things that I have tried to fix this annoying issue. (I'm slowly going crazy over this).
Tried:
- Every possible combination in the power settings of my Bios which Includes:
ErP on\off
PEG ASPM
PCH APSM
DMI ASPM
Power loading on\off
Adding:
RebootWatchdogSec=5s
ShutdownWatchdogSec=5s
DefaultTimeoutStartSec=30s
DefaultTimeoutStopSec=10s
DefaultTimeoutAbortSec=10s
To my systemd user.conf
Removed:
"nowatchdog" from my grub config since it was changed on the wiki.
After this I went to full elimination method.
I tried using Windows for a week.
Not one freeze on shutdown.
I thought maybe The Windows Nvme Is doing something funky.
So I removed it.
Still freezes.
Me using Windows successfully logically rolls out any hardware issue that might be occurring. Note the word ISSUE not incompatibility.
I reset the CMOS. And replaced the battery just because why not.
Still freezes.
Rolled back the Bios.
Still freezes.
Disconnected the Cd R/W Drive
Still freezes.
Tried using KDE
Still freezes
I got so mad that I even opened up the PC. And started disconnecting drives from Sata and Power while the PC was in this frozen state to see if any of them would allow the PC to shut down after disconnecting it.
After disconnecting all the drive. It was still going. At that point I told myself F it and disconnected the main 24pin connector (While the PC was still running the fans and the GPU light.) It went as expected the PC shut down.
I reconnected the 24Pin connector while the PSU was still under load. The fans did not come on after that. Until I started the PC back up with the power button. (I expected that)
While I'm here I'll leave the latest journal from a freeze. For anyone interested.
journalctl
The thing that I hate the most about this whole mess is that the journal ends with this:
Mar 07 01:01:44 IkorusPC systemd[1]: Reached target System Shutdown.
Mar 07 01:01:44 IkorusPC systemd[1]: Reached target Late Shutdown Services.
Mar 07 01:01:44 IkorusPC systemd[1]: systemd-poweroff.service: Deactivated successfully.
Mar 07 01:01:44 IkorusPC systemd[1]: Finished System Power Off.
Mar 07 01:01:44 IkorusPC systemd[1]: Reached target System Power Off.
Mar 07 01:01:44 IkorusPC systemd[1]: Shutting down.
Mar 07 01:01:44 IkorusPC systemd-shutdown[1]: Syncing filesystems and block devices.
Mar 07 01:01:44 IkorusPC systemd-shutdown[1]: Sending SIGTERM to remaining processes...
Mar 07 01:01:44 IkorusPC systemd-journald[378]: Received SIGTERM from PID 1 (systemd-shutdow).
Mar 07 01:01:44 IkorusPC systemd-journald[378]: Journal stopped
Indicating that all the things shut down correctly even the journal stops.
The funniest thing is that I can even hear and feel that the HDDs aren't spinning and have stopped.
Last thing I will try. Is to grab this older 240Gb SSD install Arch on it with something like KDE and see if that solves the issue.
I'll update on this.
Offline
Thank you for getting back on this in so much detail, I really appreciate it. I think we are on the same page with how annoying this power off / shutdown issue is. I've had some success with the following kernel parameters:
GRUB_CMDLINE_LINUX_DEFAULT="loglevel=3 dis_ucode_ldr nowatchdog sysrq_always_enabled=1 acpi=force acpi_rev_override acpi_osi=! acpi_osi=\"Windows 2020\""
I'm not sure if they are all needed, but the intended aim is to trick the BIOS that I'm still on Windows. See the links to kernel parameters and Windows OSI strings for more info. Now I get a clean shut down most of the time. The only re-occurrence was when Discord crashed on closing the app.
Offline
Disabled VT-d in the bios, which is enabled by default, and have had no further shutdown issues since. The idea comes from this topic: https://bbs.archlinux.org/viewtopic.php?id=297688
Offline
Anyways, it's about time I got back to you with some new info. I tried many things to fix the issue and seem to have succeeded. Although I'm not done yet.
First of all, I would like to get a confirmation from @R290 on if disabling VT-d really fixed the issue for him. For me, it didn't help. (It's not important. I'm just curious now that he has a few weeks to see if the problem reoccurs.)
Again I'm going to list the things I tried. But this time I have a solution.
1. I went and bought another PSU and tried swapping that for my current one. That didn't help.
2. I tried disabling the VT-d. Which didn't work for me.
Disabled VT-d in the bios, which is enabled by default, and have had no further shutdown issues since. The idea comes from this topic: https://bbs.archlinux.org/viewtopic.php?id=297688
3. I tried the acpi kernel line setting above. With many different Windows versions tested. That didn't work.
Thank you for getting back on this in so much detail, I really appreciate it. I think we are on the same page with how annoying this power off / shutdown issue is. I've had some success with the following kernel parameters:
GRUB_CMDLINE_LINUX_DEFAULT="loglevel=3 dis_ucode_ldr nowatchdog sysrq_always_enabled=1 acpi=force acpi_rev_override acpi_osi=! acpi_osi=\"Windows 2020\""
4. Finally. I got a great idea to go back to the beginning. I remembered that the week before I swapped out my PSU and GPU, I did a BIOS update. So I went back to the BIOS page for my motherboard and saw there was a description for each update that I just skimmed over. After I read it more closely, I noticed that a few settings were switched off by default during the jump from version to version. Mainly, in the F5 they changed the PTT status to enabled by default. Mind you, I went from F2, which isn't even listed, to F7. Now in hindsight, it seems stupid of me that I didn't think of this sooner. But oh well. Anyway, these are the BIOS settings I have switched to get my machine to shut down properly.
VT-d [Disabled]----->[Enabled]
Race To Halt [Auto]----->[Disabled]
Intel(R) Speed Shift Technology [Enabled]----->[disabled]
Native ASPM [Disabled]----->[Auto]
Intel Platform Trust Technology [Enabled]----->[disabled]
Now I'm gonna go and try turning each BIOS setting on 1 by 1 until I find the culprit. If I change all the setting back and the shutdown problem doesn't happen. That would mean the problem was fixed by a system update. Finally after that one last reply listing all the changes that were made from default BIOS/Kernel Parameters/Mkinit hooks that were needed. It might help someone in the future.
Offline
The fix with disabling VT-d was indeed temporary and did not really fix the issue. I even briefly switched to Debian to see if a much older kernel version would change anything. It did not.
The ACPI kernel line settings were also a dead end, similar to what you found out. I still have them enabled, but I think I can remove them by now since they're not doing anything useful.
With respect to the BIOS settings, I went on a similar exploration. Since I did not look at the changelog of the BIOS versions, I just disabled everything that did not seem very useful. Things like Intel Platform Trust Technology (PTT), Software Guard Extensions (SGX), XHCI hand-off, Port 60/64 emulation, Platform Power Manager (ASPM), ErP, Power Loading, RC6 (Render Standby), Internal Graphics, Wake on LAN, IOAPIC 24-119 Entries, SATA controller, Serial Port, Parallel Port, the list goes on. This did yield me some erros in journalctl:
Apr 02 18:55:21 arch kernel: x86/cpu: SGX disabled or unsupported by BIOS.
Apr 02 18:55:21 arch kernel: ERROR: Unable to locate IOAPIC for GSI 61
Apr 02 18:55:21 arch kernel: ERROR: Unable to locate IOAPIC for GSI 63
Apr 02 18:55:24 arch kernel: Serial bus multi instantiate pseudo device driver INT3515:01: error -EINVAL: IRQ index 0 not found
Apr 02 18:55:24 arch kernel: Serial bus multi instantiate pseudo device driver INT3515:01: error -EINVAL: Error requesting irq at index 0
Apr 02 18:55:24 arch kernel: Serial bus multi instantiate pseudo device driver INT3515:01: probe with driver Serial bus multi instantiate pseudo device driver failed with error -22
Althoug these seem harmless. Based on the overlap between your list and my list, it seems that most likely culprit is Intel Platform Trust Technology. We've both verified it's not VT-D nor ASPM and I have Race To Halt and Intel Speed Shift Technology still set to enabled / auto while being able to shutdown cleanly for the past week or so.
Last edited by R290 (2025-04-02 17:18:48)
Offline
I've only skimmed through the thread but
Feb 01 10:34:58 IkorusPC kernel: DMI: Gigabyte Technology Co., Ltd. B460MDS3H/B460M DS3H, BIOS F7 12/20/2023
Feb 01 10:34:58 IkorusPC kernel: acpi PNP0A08:00: FADT indicates ASPM is unsupported, using BIOS configuration
https://bbs.archlinux.org/viewtopic.php … 6#p2231896 suggested that for gigabyte, there's a BIOS configuration to hand ASPM control to the OS called "native aspm"
Try to enable that, otherwise "pcie_aspm=off" doesn't have any effect anyway.
Can you btw. properly reboot or does the system hang there as well?
Offline
Ok. I tested all the options and can now confirm. The issue was caused by Intel PTT (Intel Platform Trust Technology). The fix was to disable this in BIOS.
Once again I give thanks to everyone.
Marking the thread as solved.
Offline
Thanks to you as well for following up on this! I'm happy that a solution has been found for this annoying issue.
Offline