Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

obelisk · 2023-09-30 09:38:22

Hello,
I replaced my 1050Ti against AMD Radeon 7800 XT (Red Devil AMD Radeon™ RX 7800 XT 16GB GDDR6 RADEON™ RX 7800 XT, Productname: "RX 7800 XT 16G-E/OC").

I have the problem that I cannot turn off the PC in normal way. The PS restarts always again. It ends in a reboot loop in the point, where normally the PC goes off.

Mainboard: ASrock B550M Steel Legend
CPU: AMD Ryzen 7 5800X
GPU: AMD Radeon 7800 XT, Device 747e

$ uname -r
6.5.5-arch1-1
$ lspci | grep -i vga
08:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Device 747e (rev c8)

last journalctl log after "shutwon -h now":

$journalctl --boot=-1
#...many lines above...
.....
Sep 30 11:24:16 tiberius systemd[1]: Unmounted /home.
Sep 30 11:24:16 tiberius systemd[1]: Reached target Unmount All Filesystems.
Sep 30 11:24:16 tiberius systemd[1]: systemd-fsck@dev-disk-by\x2duuid-28f72a17\x2d4101\x2d4ebb\x2db130\x2deb3a7reec4e2.service: Deactivated successfully.
Sep 30 11:24:16 tiberius systemd[1]: Stopped File System Check on /dev/disk/by-uuid/28f72ac7-4101-4e4d-b1a0-eb3a761bc4ed.
Sep 30 11:24:16 tiberius systemd[1]: Removed slice Slice /system/systemd-fsck.
Sep 30 11:24:16 tiberius systemd[1]: Stopped target Preparation for Local File Systems.
Sep 30 11:24:16 tiberius systemd[1]: Stopping Monitoring of LVM2 mirrors, snapshots etc. using dmeventd or progress polling...
Sep 30 11:24:16 tiberius systemd[1]: systemd-remount-fs.service: Deactivated successfully.
Sep 30 11:24:16 tiberius systemd[1]: Stopped Remount Root and Kernel File Systems.
Sep 30 11:24:16 tiberius systemd[1]: systemd-tmpfiles-setup-dev.service: Deactivated successfully.
Sep 30 11:24:16 tiberius systemd[1]: Stopped Create Static Device Nodes in /dev.
Sep 30 11:24:16 tiberius systemd[1]: systemd-tmpfiles-setup-dev-early.service: Deactivated successfully.
Sep 30 11:24:16 tiberius systemd[1]: Stopped Create Static Device Nodes in /dev gracefully.
Sep 30 11:24:16 tiberius systemd[1]: lvm2-monitor.service: Deactivated successfully.
Sep 30 11:24:16 tiberius systemd[1]: Stopped Monitoring of LVM2 mirrors, snapshots etc. using dmeventd or progress polling.
Sep 30 11:24:16 tiberius systemd[1]: Reached target System Shutdown.
Sep 30 11:24:16 tiberius systemd[1]: Reached target Late Shutdown Services.
Sep 30 11:24:16 tiberius systemd[1]: systemd-poweroff.service: Deactivated successfully.
Sep 30 11:24:16 tiberius systemd[1]: Finished System Power Off.
Sep 30 11:24:16 tiberius systemd[1]: Reached target System Power Off.
Sep 30 11:24:16 tiberius systemd[1]: Shutting down.
Sep 30 11:24:16 tiberius systemd-shutdown[1]: Syncing filesystems and block devices.
Sep 30 11:24:16 tiberius systemd-shutdown[1]: Sending SIGTERM to remaining processes...
Sep 30 11:24:16 tiberius systemd-journald[381]: Received SIGTERM from PID 1 (systemd-shutdow).
Sep 30 11:24:16 tiberius systemd-journald[381]: Journal stopped
lines 2663-2716/2716 (END)

journalctl log of boot looks normal.
Kernel parameter are (systemd-boot)

$ cat /boot/loader/entries/arch.conf 
title   Archlinux on SAM
linux   /vmlinuz-linux
initrd  /amd-ucode.img
initrd  /initramfs-linux.img
options root=UUID=d9******-****-****-****-****e0  rw loglevel=2 amd_pstate=passive lang=de locale=de_DE.UTF-8

what is installed:

$ pacman -Qs radeon
local/amdgpu-pro-oglp 23.20_1646757-1 (Radeon_Software_for_Linux)
    AMDGPU Pro OpenGL driver
local/amdgpuinfo-git 0.2.0.146.1f0e093-1
    Get information from AMD Radeon GPUs
local/amf-amdgpu-pro 23.20_1646757-1 (Radeon_Software_for_Linux)
    AMDGPU Pro Advanced Multimedia Framework
local/hsakmt-roct 5.6.1-1
    Radeon Open Compute Thunk Interface
local/lib32-amdgpu-pro-oglp 23.20_1646757-1 (Radeon_Software_for_Linux)
    AMDGPU Pro OpenGL driver (32-bit)
local/lib32-vulkan-amdgpu-pro 23.20_1646757-1 (Radeon_Software_for_Linux)
    AMDGPU Pro Vulkan driver (32-bit)
local/lib32-vulkan-radeon 1:23.1.8-1
    Radeon's Vulkan mesa driver (32-bit)
local/radeontop 1.4-2
    View GPU utilization for total activity percent and individual blocks
local/vulkan-amdgpu-pro 23.20_1646757-1 (Radeon_Software_for_Linux)
    AMDGPU Pro Vulkan driver
local/vulkan-radeon 1:23.1.8-1
    Radeon's Vulkan mesa driver

$ pacman -Qs amd
local/amd-ucode 20230804.7be2766d-2
    Microcode update image for AMD CPUs
local/amd-vulkan-prefixes 1-1
    Select needed vulkan implementation with vk_radv, vk_amdvlk or vk_pro prefix
local/amdfand-bin 1.0.13-4
    AMDGPU fan speed management tool
local/amdgpu-pro-oglp 23.20_1646757-1 (Radeon_Software_for_Linux)
    AMDGPU Pro OpenGL driver
local/amdgpu_top 0.2.0-1
    Tool that shows AMD GPU utilization
local/amdgpuinfo-git 0.2.0.146.1f0e093-1
    Get information from AMD Radeon GPUs
local/amdguid-wayland-bin 1.0.11-2
    AMDGPU temperature and fan speed monitoring tool
local/amdvlk 2023.Q3.3-1
    AMD's standalone Vulkan driver
local/amf-amdgpu-pro 23.20_1646757-1 (Radeon_Software_for_Linux)
    AMDGPU Pro Advanced Multimedia Framework
local/amf-headers 1.4.30-1
    Header files for AMD Advanced Media Framework
local/lib32-amdgpu-pro-oglp 23.20_1646757-1 (Radeon_Software_for_Linux)
    AMDGPU Pro OpenGL driver (32-bit)
local/lib32-amdvlk 2023.Q3.3-1
    AMD's standalone Vulkan driver
local/lib32-vulkan-amdgpu-pro 23.20_1646757-1 (Radeon_Software_for_Linux)
    AMDGPU Pro Vulkan driver (32-bit)
local/nvtop 3.0.2-1
    GPUs process monitoring for AMD, Intel and NVIDIA
local/texlive-latexextra 2023.66594-19 (texlive)
    TeX Live - LaTeX additional packages
local/vulkan-amdgpu-pro 23.20_1646757-1 (Radeon_Software_for_Linux)
    AMDGPU Pro Vulkan driver
local/xf86-video-amdgpu 23.0.0-1 (xorg-drivers)
    X.org amdgpu video driver
local/zenmonitor3-git 90.a09f0b2-1
    Zenmonitor3 is monitoring software for AMD Zen-based CPUs, now with Zen 3 support!
local/zenpower-dkms-git 46.5e2f56f-1
    Linux kernel driver for reading sensors for AMD Zen family CPUs
local/zenpower3-dkms 0.2.0-3
    Linux kernel driver for reading sensors for AMD Zen family CPUs

-I found with searching to add xhci_hcd.quirks=270336 into the kernel parameter.... But it brings nothing. Still not able to normal power off the PC.
-I tried also to disconnect all USB cables from the Mainboard... same result, when I press the power button on the PC to poweroff the PC.

I'm not satisfied with the thing, to turn off 230V at the moment where the PC normally must poweroff.
So, have anybody an Idea how to solve , that I can normal poweroff the PC?

Last edited by obelisk (2023-09-30 09:39:16)

evilch3n · 2023-10-02 11:27:09

Same problem but different hardware here, didn't know it would be the GPU.. damn

Mainboard: MSI B550 PRO
CPU: AMD Ryzen 7 5800x3d
GPU: Sapphire Nitro AMD Radeon 7800 XT (OS Cant read the name right now)
Kernel: 6.5.5

(The OS cant read the GPU Name right now, i guess its not fully implemented yet?)

obelisk · 2023-10-02 14:41:04

I tested the 7800 XT on a aother PC, with Dual Boot. Archlinux with kernel 6.5.5 had same problem. Shutdown from Windows 10 is working.
So yes, I think the GPU is not fully implemented yet.

I tied with default uefi-bios settings, installed linux-firmware-git, AMD drivers manually 23.20 installed with debtab....Everything has not worked.
I thin, because amdgpu is compiled in the kernel, it takes some time till new kernel will solve this. But I found no solution for the shutdown problem. I searched nearly the whole weekend for solution. So I sent the card today back to the dealer, mainly because of the shutdown problem and some other points.

In the meantime I had everything running, like e.g.:

- Games Vulkan with amdvlk (manual installed the new v-2023.Q3.3)

- Games Vulkan with RADV

- Games Vulkan with amdgpu_pro

- Video encoding (only amdgpu_pro) , with h264_amf, hevc_amf and also av1_amf (mpv cannot forward while playing, but with "mpv -vd libaom-av1 AV1_FILE.mkv" it is possible). Video encoding with the AMD was cool. Nvidia has a (license?) limitation on parallel files which can be encoded at same time. With the Radeon started 14 in parallel, then I aborted the test. Makes no sense to encode more in parallel, speed goes down. But I have no speed reference to a actual Nvidia card.

- 10-bit colour worked nearly fine. Some applications like tvbrowser (java error reports) or steam won't start with 10-bit colour. And Vulkan with amdvlk and amdgpu_pro have colour problem: everything is everything is colorful. For activating 10-bi colour I created a file:

Section "Screen"
$ cat /etc/X11/xorg.conf.d/22-screen.conf
Section "Screen"
    Identifier     "Screen0"
    #DefaultDepth    24	#8-bit Color
    DefaultDepth    30	#10-Bit Color   
EndSection

And then it can be tested easy with

vk_radv vkcube
vk_amdvlk vkcube
vk_pro vkcube

- I had Freesync (with a Freesync compatible monitor) also working good:

#Show avaiable parameters of module amdgpu:
modinfo amdgpu
#set parameter:
cat /etc/modprobe.d/amdgpu.conf 
options freesync_video=1
#check loaded kernel parameters (after reboot):
systool -v -m amdgpu | grep -i freesync_video
#another testing:
xrandr --props | grep vrr_capable
#or
vrrtest

seth · 2023-10-02 15:21:41

Can you either of

systemctl poweroff --force --force
systemctl poweroff --force

https://bbs.archlinux.org/viewtopic.php?id=289190

DinoNugget1337 · 2023-10-02 16:02:05

Hey, as you can see from the thread seth linked I have the same issue with my RX 7800 XT.

obelisk · 2023-10-03 07:39:02

@ seth: inconveniently, I sent the card back to the dealer yesterday. I have no card to test anymore. Too bad, if I had known that a day before, I would have tried it with the force.

@ DinoNugget1337: Yes, looks like same issue. I found on another site somebody who had also same issue, which has solved it with blacklist the watchdog module "sp5100_tco". But my PC hadn't this watchdog, so this also didn't help me. I'm not sure if its allowed to add url's to external websites, I hope I do not get into trouble, the website is: https://forum.endeavouros.com/t/new-ins … tart/45384

I hope that somebody find a solution..even if I no longer have a 7800 XT card, I would like to know the solution is.

DinoNugget1337 · 2023-10-03 09:17:41

obelisk wrote:

@ DinoNugget1337: Yes, looks like same issue. I found on another site somebody who had also same issue, which has solved it with blacklist the watchdog module "sp5100_tco". But my PC hadn't this watchdog, so this also didn't help me

Yeah, same for me.

evilch3n · 2023-10-03 12:30:24

seth wrote:

Can you either of
systemctl poweroff --force --force
systemctl poweroff --force
https://bbs.archlinux.org/viewtopic.php?id=289190

hi, thanks for jumping in - sadly both of those commands did not help. system powers off -> restarts

edit:

> blacklist the watchdog module "sp5100_tco"

what exactly does "sp5100_tco" do? is it save to disable?

Last edited by evilch3n (2023-10-03 12:32:27)

seth · 2023-10-03 12:43:55

https://wiki.archlinux.org/title/Improv … #Watchdogs

lsmod | grep -i tco

to see whether and which is loaded at all.

Can you poweroff the system when booting the multi-user.target (2nd link below) along "nomodeset"?
https://wiki.archlinux.org/title/Kernel_parameters

evilch3n · 2023-10-03 13:29:45

seth wrote:

https://wiki.archlinux.org/title/Improv … #Watchdogs
lsmod | grep -i tco
to see whether and which is loaded at all.
Can you poweroff the system when booting the multi-user.target (2nd link below) along "nomodeset"?
https://wiki.archlinux.org/title/Kernel_parameters

Added

nomodeset

to kernel-params, system did shutdown properly 3x in a row **BUT** now no 3D programs function, steam wont load and the monitor-refreshrate is set to 60 (was 144hz before) and cant be changed.

When launching steam journalctl outputs:

Could not apply discrete GPU environment, switcheroo-control not available

btw: big thanks for your trying to help so far

Last edited by evilch3n (2023-10-03 13:32:05)

seth · 2023-10-03 13:32:53

**BUT** now no 3D programs function

That's completely expectable, I just wanted to confirm a pattern in the other thread (and pinning this on the amdgpu chip)

TcCoy · 2023-10-05 08:48:52

Hi all,

Thanks for this thread that I follow with great interest. I have the exact same issue with also a RX 7800 XT, from XFX, on a newly built PC two days ago.

After suspecting a Wake-On-Lan issue on the motherboard and trying BIOS tweaks and kernel parameters as indicated in the wiki, I noticed the reboot issue happened when the HDMI cable used to connect the PC screen was plugged to the GPU but not if it was plugged to the HDMI port of the motherboard. I believed then that something was wrong with the cable management in the PC case, and I improved it to avoid any contacts with the GPU, without effects.

Finally, the issue disappeared when connecting the HDMI cable to a TV instead of the PC screen, which seems like sorcery to me . I intend to buy a cable to connect the GPU to the PC screen via a DisplayPort instead, but I am not confident in the result.

Are there any logs or further information that I can provide to help ?

Last edited by TcCoy (2023-10-05 08:49:17)

evilch3n · 2023-10-06 15:08:36

seth wrote:

**BUT** now no 3D programs function
That's completely expectable, I just wanted to confirm a pattern in the other thread (and pinning this on the amdgpu chip)

ahh got it! thanks

Are there any logs or further information that I can provide to help ?

Kernel version would be interesting, because i read someone having the same issue on another forum and they said the mainline kernel (which is in rc4 right now) did not have the wake-up problem + the 7800 got identified correctly.

Since nothing worked so far my/our only hope is a kernel update/fix?

Sardok77 · 2023-10-06 15:23:11

Similar problem as obelisk, evilch3n and DinoNugget1337 but different hardware:

Mainboard: MSI MPG B550 Gaming Edge WIFI
CPU: AMD Ryzen 7 5800x
GPU: Sapphire Nitro AMD Radeon 7800 XT
Kernel: 6.5.5
lastet linux-firmware-git from the AUR
all Packages updated with pacman

However, I have the following discrepancy. When the PC just reboots after shutdown, I get the following mce error message. (My brother has the same hardware with the same problem since the installation of the new graphics card):

$journalctl -b0
#...many lines above...
.....
Okt 06 16:17:53 Panther kernel: mce: [Hardware Error]: CPU 0: Machine Check: 0 Bank 5: bea0000001000108
Okt 06 16:17:53 Panther kernel: microcode: CPU15: patch_level=0x0a201025
Okt 06 16:17:53 Panther kernel: mce: [Hardware Error]: TSC 0 ADDR ffffff978103a0 
Okt 06 16:17:53 Panther kernel: fbcon: Taking over console
Okt 06 16:17:53 Panther kernel: MISC d012000100000000 SYND 4d000000 IPID 500b000000000 
Okt 06 16:17:53 Panther kernel: mce: [Hardware Error]: PROCESSOR 2:a20f10 TIME 1696601867 SOCKET 0 APIC 0 microcode a201025
#...many lines after...

However, I was able to solve the problem after many unsuccessful attempts:

Unlike DinoNugget1337, my problem was fixed with the linux-mainline 6.6rc4-1 kernel from the AUR. The PC now shuts down without having to press the power button. In addition, there is no longer a mce error message when rebooting.

evilch3n i would advise you to try this kernel. Your hardware is very similar to mine. Maybe it will fix the problem for you too.

Last edited by Sardok77 (2023-10-06 15:28:54)

DinoNugget1337 · 2023-10-06 17:47:20

Also just tried again with the newest linux-firmware-git and linux-mainline 6.6rc4-1 and while the 7800XT is now correctly displayed for me as well instead of just "AMD ATI 03:00.0 Device 747e" the problem is still the same. At least Linux Mint is working for me (linux-firmware-git and 6.5.5 kernel) so I feel like I have to use that one until it's hopefully fixed soon.

seth · 2023-10-06 19:58:08

Do you reboot to MCEs w/ 6.6rc4?
Does 6.6rc4 still fail to shutdown from the multi-user.target?

Sardok77 · 2023-10-07 12:59:35

seth wrote:

Do you reboot to MCEs w/ 6.6rc4?

No. With kernel 6.6rc4 the mce errors are gone.

seth wrote:

Does 6.6rc4 still fail to shutdown from the multi-user.target?

No. With kernel 6.6rc4 the PC shuts down normally.

With one exception. I don't know why, but with kernel 6.5.5 and kernel 6.6rc4, Xfce's whisker menu causes a segfault on shutdown. I could not find any negative effects except for this coredump that is created.

Okt 07 10:53:34 Panther kernel: panel-7-whisker[1475]: segfault at 5612adf95088 ip 00007fd29f445f41 sp 00007ffc8964ae98 error 4 in libgobject-2.0.so.0.7800.0[7fd29f41a000+35000] likely on CPU 4 (core 4, socket 0)
Okt 07 10:53:34 Panther kernel: Code: c0 0f 95 c0 48 83 c4 08 0f b6 c0 c3 66 66 2e 0f 1f 84 00 00 00 00 00 66 90 f3 0f 1e fa 48 85 ff 74 47 48 8b 07 48 85 c0 74 3f <48> 8b 00 48 3d fc 03 00 00 77 2c 48 8d 15 ed 61 02 00 48 c1 e8 02
Okt 07 10:53:34 Panther systemd[1]: Created slice Slice /system/systemd-coredump.
Okt 07 10:53:34 Panther systemd[1]: Started Process Core Dump (PID 5064/UID 0).
Okt 07 10:53:34 Panther systemd-coredump[5068]: Process 1475 (panel-7-whisker) of user 1000 dumped core.

Kernel 6.5.5 always reboots after a poweroff. On reboot mce-errors appear in the log. Also when I reboot with this kernel I get mce-errors after the reboot.

Logs from Kernel 6.5.5:
Log after poweroff followed by reboot. The mce errors are contained there.
journalctl -b -0

Log before the reboot. Here I wanted to turn off the PC. Instead the PC is restarted.
journalctl -b -1

Logs from Kernel 6.6rc4:
Log from current boot. No mce-errors present:
journalctl -b -0

Here, the PC is shut down correctly. Unfortunately, as with kernel 6.5.5, there is also a segfault from the XFCE whisker menu on shutdown.
journalctl -b -1

seth · 2023-10-07 13:50:47

I actually meant to address DinoNugget1337 (to rule out that he encounters a secondary issue when shutting down from the session)
About whisker: https://wiki.archlinux.org/title/Core_d … _core_dump
You might have to https://wiki.archlinux.org/title/Debugg … g_packages to get a backtrace more meaningful than "crashed somewhere"

DinoNugget1337 · 2023-10-07 15:13:21

seth wrote:

Do you reboot to MCEs w/ 6.6rc4?
Does 6.6rc4 still fail to shutdown from the multi-user.target?

Shutdown from multi-user.target still fails with that, yes. I'm not sure what MCEs are but if a log helps here is one: https://0x0.st/HWGF.txt

Last edited by DinoNugget1337 (2023-10-07 15:13:52)

evilch3n · 2023-10-07 16:02:13

DinoNugget1337 wrote:

seth wrote:
Do you reboot to MCEs w/ 6.6rc4?
Does 6.6rc4 still fail to shutdown from the multi-user.target?
Shutdown from multi-user.target still fails with that, yes. I'm not sure what MCEs are but if a log helps here is one: https://0x0.st/HWGF.txt

maybe you are suffering from additional or other issues here, new mainline seems to help - did you test the kernel tweak suggestions for WOL issues from the wiki with the mainline-kernel? maybe those help now

https://wiki.archlinux.org/title/Wake-o … nel_quirks

DinoNugget1337 · 2023-10-07 16:57:34

evilch3n wrote:

maybe you are suffering from additional or other issues here, new mainline seems to help - did you test the kernel tweak suggestions for WOL issues from the wiki with the mainline-kernel? maybe those help now
https://wiki.archlinux.org/title/Wake-o … nel_quirks

WOL was indeed turned on in the BIOS but turning that off and setting the suggested kernel parameter as well didn't help either. If the mainline kernel helps others with this problem but didn't fix mine then, yes, I think there are other factors for me that cause these issues. Are there maybe other important BIOS settings that should be disabled that could cause this? The only things I changed in there were turning on XMP and disabling fast boot.

seth · 2023-10-07 17:35:49

https://en.wikipedia.org/wiki/Machine-check_exception - and there're none registered.
Ftr, you didn't stop at nor reboot from the multi-user.target in the boot you posted.

Oct 07 17:07:44 stoepsel-pc /usr/lib/gdm-x-session[1135]: cinnamon-session[1135]: WARNING: t+26.90755s: Requesting system restart...
Oct 07 17:07:44 stoepsel-pc cinnamon-session[1135]: WARNING: t+26.90755s: Requesting system restart...
Oct 07 17:07:44 stoepsel-pc /usr/lib/gdm-x-session[1135]: cinnamon-session[1135]: WARNING: t+26.90759s: Attempting to restart using systemd...
Oct 07 17:07:44 stoepsel-pc cinnamon-session[1135]: WARNING: t+26.90759s: Attempting to restart using systemd...
Oct 07 17:07:44 stoepsel-pc systemd-logind[485]: The system will reboot now!
Oct 07 17:07:44 stoepsel-pc systemd-logind[485]: System is rebooting.

XMP can lead to general instabilities, but your situation seems rahter consistent and hinge on newer kernels.

Edit: One more thing you could try and wrt XMP, pass

amdgpu.gartsize=0 amdgpu.gttsize=0 amdgpu.smu_memory_pool_size=0

to the kernel.

Last edited by seth (2023-10-07 19:30:10)

DinoNugget1337 · 2023-10-08 10:22:28

amdgpu.gartsize=0 amdgpu.gttsize=0 amdgpu.smu_memory_pool_size=0

Issue still persists with these as well. Tbh I'm a bit frustrated at this point but I kinda should've known better buying new hardware on launch I think I'm gonna stay on mint for now and see if updating in a month or so resolves it on arch. Still, appreciate all the help until now tho so big thanks for that.

TcCoy · 2023-10-08 20:43:37

evilch3n wrote:

Kernel version would be interesting, because i read someone having the same issue on another forum and they said the mainline kernel (which is in rc4 right now) did not have the wake-up problem + the 7800 got identified correctly.
Since nothing worked so far my/our only hope is a kernel update/fix?

Sorry for responding so late. The behavior I described was on kernel 6.5.5 (same as Obelisk's first post), while the output of :

lspci | grep -i vga

was also the same as Obelisk's.

In fact, after further testing, it appeared that my PC mostly shuts down properly, but it still occasionally reboots.

I am now with kernel :

6.5.6-arch2-1

with :

lspci | grep -i vga
03:00.0 VGA compatible controller: Advanced Micro Devices, Inc. [AMD/ATI] Navi 32 [Radeon RX 7700 XT / 7800 XT] (rev c8)

which is a step forward I think, but poweroff remains unsure. I will dig into logs of both situations as soon as I have a little more time (this week I hope).

DinoNugget1337 wrote:

I kinda should've known better buying new hardware on launch

Indeed, but I believe it could have been worse (suspend and hibernate can certainly help waiting for further updates).

spiffeeroo · 2023-10-10 14:00:38

I also have a 7800 XT and AMD Ryzen 5xxx CPU.

On kernel 6.5.5/6.5.6, going from cold boot -> reboot (mce error) -> shutdown always resulted in failed shutdown and reboot. However going from cold boot -> shutdown worked normally.

Kernel mainline 6.6-rc4/rc5 fixes the mce and shutdown problem. No mce error if cold booting (and subsequently rebooting) into kernel 6.6-rc4.

Arch Linux

#1 2023-09-30 09:38:22

Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#2 2023-10-02 11:27:09

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#3 2023-10-02 14:41:04

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#4 2023-10-02 15:21:41

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#5 2023-10-02 16:02:05

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#6 2023-10-03 07:39:02

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#7 2023-10-03 09:17:41

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#8 2023-10-03 12:30:24

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#9 2023-10-03 12:43:55

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#10 2023-10-03 13:29:45

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#11 2023-10-03 13:32:53

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#12 2023-10-05 08:48:52

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#13 2023-10-06 15:08:36

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#14 2023-10-06 15:23:11

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#15 2023-10-06 17:47:20

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#16 2023-10-06 19:58:08

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#17 2023-10-07 12:59:35

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#18 2023-10-07 13:50:47

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#19 2023-10-07 15:13:21

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#20 2023-10-07 16:02:13

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#21 2023-10-07 16:57:34

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#22 2023-10-07 17:35:49

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#23 2023-10-08 10:22:28

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#24 2023-10-08 20:43:37

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

#25 2023-10-10 14:00:38

Re: Cannot Shutdown/poweroff with AMD Radeon 7800 XT, poweroff>rebootloop

Board footer