System crashes directly during or after boot

chrislg · 2025-02-08 14:49:41

My system:
AMD Ryzen 9 7950X3D
Corsair iCUE H100x RGB ELITE 47.73 CFM Liquid CPU Cooler
Gigabyte B650 GAMING X AX V2 ATX AM5 Motherboard
Corsair Vengeance 32 GB (2 x 16 GB) DDR5-6000 CL30 Memory
Samsung 990 Pro 2 TB M.2-2280 PCIe 4.0 X4 NVME Solid State Drive
Corsair RM850e (2023) 850 W 80+ Gold Certified Fully Modular ATX Power Supply
I also have an ASUS Phoenix Radeon RX 550 and have the latest F32g BIOS Update

the problem arised suddenly under 6.12.10 with KDE Plasma 6.3 beta 2 although the system runned smoothly for many weeks.
I firstly saw some KWin crashes but then it started to crash during the boot process. It also crashes during the boot of a recovery Arch ISO or even after FreeBSD installation. I reinstalled Arch with the latest stable kde plasma version and 6.13.1 kernel and the problem remains. I do not know if it is a software issue but I suspect hardware since it appears even with FreeBSD ...

I attach my dmesg and journalctl outputs.

https://pastebin.com/tMdKchRj
https://pastebin.com/uDS9gzd6

Last edited by chrislg (2025-02-08 14:56:27)

chrislg · 2025-02-09 17:18:16

Hey guys, I think I solved the problem. I downgraded the BIOS version of my B650 motherboard.

cryptearth · 2025-02-09 17:41:48

chrislg wrote:

Hey guys, I think I solved the problem. I downgraded the BIOS version of my B650 motherboard.

can you provide specific versions?

chrislg · 2025-02-09 18:43:44

From F32g to F32f. I had no problems for 3 hours (usually the system freezed and crashed 5-10 minutes after boot). Now I got 2 more crashes ... I really do not know.
I got many

Feb 09 19:35:03 archlinux kwin_wayland[27143]: kwin_scene_opengl: 0x2: GL_INVALID_OPERATION in glDrawBuffers(unsupported buffer GL_BACK_LEFT)

errors right before the crash.

I am not sure that the problem is timely connected to my BIOS update. I updated the BIOS on 4th of January and problems appeared late January...

Last edited by chrislg (2025-02-09 18:44:35)

cryptearth · 2025-02-09 19:08:47

I'm no expert but the archiso boot shows a crash of the wifi stuff (don't understand why people keep buying these mobos) - have you tried to blacklist the wifi driver or disable it in bios?

chrislg · 2025-02-11 14:39:21

The crash happens most of the time when I watch YouTube videos. The screen freezes and the audio does a 1 sec loop. The only way to restore the crash is to shut down from the power button

chrislg · 2025-02-12 09:02:05

Do you think it is a kernel or hardware issue? Any ideas?

seth · 2025-02-12 09:43:08

The screen freezes and the audio does a 1 sec loop.

is typically a kernel panic - whether a kernel bug or in response to some HW failure is impossible to tell w/o further data on the problem.

The posted journal is from the installation iso - that didn't crash, did it?
It shows a warning from the broadcom WL driver

If you cannot get away from that w/ https://wiki.archlinux.org/title/Keyboa … el_(SysRq) you'll have to use https://wiki.archlinux.org/title/Kdump to capture the crash.

There's generally https://wiki.archlinux.org/title/Ryzen#Troubleshooting

chrislg · 2025-02-12 13:39:56

The ISO did not crash, sorry I will post the logs from the installed system and I will try to run an LTS kernel

chrislg · 2025-02-12 20:22:00

I did a fresh install and booted with linux-lts. The system froze a few seconds after playing a YouTube video...

I get this error:

Feb 12 21:11:45 archlinux kernel: amdgpu 0000:0f:00.0: [drm] *ERROR* lttpr_caps phy_repeater_cnt is 0x0, forcing it to 0x80.

and this is directly before the crash in multiple crashes:

Feb 12 21:15:58 archlinux plasmashell[1277]: The cached device pixel ratio value was stale on window update.  Please file a QTBUG which explains how to reproduce.

The journalctl output: https://pastebin.com/LM0b6eZa

Last edited by chrislg (2025-02-12 20:24:39)

seth · 2025-02-12 20:56:42

How do you define "crash"?
There's no coredump etc. in that journal, do you have the same problems when logging into a fresh user account?
Do have the same issue w/ plasma on X11?

chrislg · 2025-02-12 22:14:15

I have the same issue with X11, other desktop environments eg. XFCE and after login into a fresh account the system is not even a year old (was built on May 2024)

Last edited by chrislg (2025-02-12 22:22:03)

seth · 2025-02-12 22:59:31

seth wrote:

How do you define "crash"?
There's no coredump etc. in that journal

Can you switch the VT afterwards or reboot w/ the https://wiki.archlinux.org/title/Keyboa … el_(SysRq) (which needs to be explicitly enabled first!)

chrislg · 2025-02-14 23:39:58

I get these error messages before the system crash. I can do nothing when the system freezes even with sysrq set to 1 or 244. The system can not reboot using Alt + SysRq + b or change into tty mode. I am really desperate at this point, it just sucks

Feb 15 00:21:35 archlinux systemd[939]: Reached target Bluetooth.
Feb 15 00:21:35 archlinux systemd-logind[818]: Watching system buttons on /dev/input/event1 (Power Button)
Feb 15 00:21:35 archlinux kwin_wayland[1018]: kwin_wayland_drm: failed to open drm device at ""
Feb 15 00:21:35 archlinux kwin_wayland[1018]: kwin_wayland_drm: failed to open drm device at ""
Feb 15 00:21:35 archlinux kwin_wayland[1018]: kwin_wayland_drm: failed to open drm device at ""
Feb 15 00:21:35 archlinux kwin_wayland[1018]: kwin_wayland_drm: failed to open drm device at ""
Feb 15 00:21:35 archlinux kwin_wayland[1018]: kwin_wayland_drm: failed to open drm device at ""
Feb 15 00:21:35 archlinux systemd-logind[818]: Watching system buttons on /dev/input/event0 (Power Button)
Feb 15 00:21:35 archlinux systemd-logind[818]: Watching system buttons on /dev/input/event10 (HP, Inc HyperX Alloy Core RGB Consumer Control)
Feb 15 00:21:35 archlinux systemd-logind[818]: Watching system buttons on /dev/input/event9 (HP, Inc HyperX Alloy Core RGB System Control)
Feb 15 00:21:35 archlinux systemd-logind[818]: Watching system buttons on /dev/input/event12 (HP, Inc HyperX Alloy Core RGB)
Feb 15 00:21:35 archlinux systemd-logind[818]: Watching system buttons on /dev/input/event7 (HP, Inc HyperX Alloy Core RGB)
Feb 15 00:21:36 archlinux kded6[1179]: org.kde.colorcorrectlocationupdater: Geolocator stopped
Feb 15 00:22:00 archlinux systemd[1]: Reloading D-Bus System Message Bus...
Feb 15 00:22:00 archlinux systemd[1]: Reloaded D-Bus System Message Bus.
Feb 15 00:22:00 archlinux sudo[3379]: pam_unix(sudo:session): session closed for user root
Feb 15 00:22:01 archlinux kded6[1179]: org.kde.colorcorrectlocationupdater: Geolocator stopped
Feb 15 00:22:54 archlinux kwin_wayland_wrapper[1018]: KCrash: Application 'kwin_wayland' crashing... crashRecursionCounter = 2
Feb 15 00:22:54 archlinux systemd-coredump[26123]: Process 1018 (kwin_wayland) of user 1000 terminated abnormally with signal 8/FPE, processing...
Feb 15 00:22:54 archlinux systemd[1]: Created slice Slice /system/drkonqi-coredump-processor.
Feb 15 00:22:54 archlinux systemd[1]: Created slice Slice /system/systemd-coredump.
Feb 15 00:22:54 archlinux systemd[1]: Started Process Core Dump (PID 26123/UID 0).
Feb 15 00:22:54 archlinux systemd[1]: Started Pass systemd-coredump journal entries to relevant user for potential DrKonqi handling.
Feb 15 00:22:55 archlinux (sd-parse-elf)[26127]: Could not parse number of program headers from core file: invalid `Elf' handle
Feb 15 00:22:55 archlinux (sd-parse-elf)[26127]: Could not parse number of program headers from core file: invalid `Elf' handle
Feb 15 00:22:55 archlinux (sd-parse-elf)[26127]: Could not parse number of program headers from core file: invalid `Elf' handle
Feb 15 00:22:55 archlinux systemd-coredump[26124]: Process 1018 (kwin_wayland) of user 1000 dumped core.
                                                   
                                                   Module [dso] without build-id.
                                                   Module [dso] without build-id.
                                                   Module [dso]
                                                   Module [dso] without build-id.
                                                   Module [dso]
                                                   Module libKF6BreezeIcons.so.6 without build-id.
                                                   Module [dso] without build-id.
                                                   Module [dso] without build-id.
                                                   Module [dso]
                                                   Module [dso] without build-id.
                                                   Module [dso] without build-id.
                                                   Module [dso] without build-id.
                                                   Module [dso] without build-id.
                                                   Module [dso] without build-id.
                                                   Module [dso] without build-id.
                                                   Module [dso] without build-id.
                                                   Stack trace of thread 1018:
                                                   #0  0x00007b68a12a5624 n/a (n/a + 0x0)
                                                   #1  0x00007b68a124bba0 n/a (n/a + 0x0)
                                                   #2  0x00007b68a4a182a1 n/a (n/a + 0x0)
                                                   #3  0x00007b68a124bcd0 n/a (n/a + 0x0)
                                                   #4  0x00007b68a4612201 n/a (n/a + 0x0)
                                                   #5  0x00007b68a463fd40 n/a (n/a + 0x0)
                                                   #6  0x00007b68a4638cb0 n/a (n/a + 0x0)
                                                   #7  0x00007b68a463a7eb n/a (n/a + 0x0)
                                                   #8  0x00007b68a463c509 n/a (n/a + 0x0)
                                                   #9  0x00007b68a437ae80 n/a (n/a + 0x0)
                                                   #10 0x00007b68a19b1a49 n/a (n/a + 0x0)
                                                   #11 0x00007b68a4390835 n/a (n/a + 0x0)
                                                   #12 0x00007b68a4395033 n/a (n/a + 0x0)
                                                   #13 0x00007b68a19b1a49 n/a (n/a + 0x0)
                                                   #14 0x00007b68a19baa75 n/a (n/a + 0x0)
                                                   #15 0x00007b68a19a2ef9 n/a (n/a + 0x0)
                                                   #16 0x00007b68a2aff0ca n/a (n/a + 0x0)
                                                   #17 0x00007b68a1955b00 n/a (n/a + 0x0)
                                                   #18 0x00007b68a1ad5aff n/a (n/a + 0x0)
                                                   #19 0x00007b68a1ade408 n/a (n/a + 0x0)
                                                   #20 0x00007b68a2661d93 n/a (n/a + 0x0)
                                                   #21 0x00007b68a19606a6 n/a (n/a + 0x0)
                                                   #22 0x00007b68a19591d6 n/a (n/a + 0x0)
                                                   #23 0x00006269ef6d607f n/a (n/a + 0x0)
                                                   #24 0x00007b68a1235488 n/a (n/a + 0x0)
                                                   #25 0x00007b68a123554c n/a (n/a + 0x0)
                                                   #26 0x00006269ef6dc3f5 n/a (n/a + 0x0)
                                                   ELF object binary architecture: AMD x86-64
Feb 15 00:22:55 archlinux systemd[1]: systemd-coredump@0-26123-0.service: Deactivated successfully.
Feb 15 00:22:55 archlinux systemd[1]: systemd-coredump@0-26123-0.service: Consumed 470ms CPU time, 378.8M memory peak.
Feb 15 00:22:55 archlinux plasmashell[1257]: The Wayland connection broke. Did the Wayland compositor die?
Feb 15 00:22:55 archlinux DiscoverNotifier[1574]: The Wayland connection broke. Did the Wayland compositor die?
Feb 15 00:22:55 archlinux kwin_wayland_wrapper[1107]: (EE) failed to read Wayland events: Broken pipe

Here are the journalctl logs just from today with 4-5 crashes: https://justpaste.it/h52rm

Last edited by chrislg (2025-02-15 00:02:29)

tekstryder · 2025-02-15 02:19:47

Regarding this error specifically:

(sd-parse-elf)[26127]: Could not parse number of program headers from core file: invalid `Elf' handle

I've observed various applications dumping core thusly sporadically, typically on exit, since the glibc 2.41 & friends update.

I investigated somewhat by reproducing with older kernels, rebuilding crashing packages against the latest toolchain (in a clean chroot), examining ELF data and execstacks of offending shared object files on the stack... but never came to a solid conclusion.

Oddly, the frequency of crash occurrence has slowed for me and I can rarely reproduce now. But, we've recently seen a few disparate forum threads with similar stacktraces.

The gdb backtraces I've looked at with debug symbols from my coredumps have yielded equally useless info.

In my instances, the issue has been limited to merely a variety of GTK4 apps crashing on close. I don't use KDE, but your crashing process looks far more frustrating!

2025-02-03 10:35:06-0500] [ALPM] upgraded glibc (2.40+r66+g7d4b6bcae91f-1 -> 2.41+r2+g0a7c7a3e283a-1)
2025-02-03 10:35:07-0500] [ALPM] upgraded gcc-libs (14.2.1+r134+gab884fffe3fc-2 -> 14.2.1+r730+gc061ad5a36ba-1)
2025-02-03 10:35:07-0500] [ALPM] upgraded libelf (0.192-2 -> 0.192-3)
2025-02-03 10:35:07-0500] [ALPM] upgraded binutils (2.43_1+r171+g01da089627be-1 -> 2.43_1+r186+g61f8adadd6db-1)
2025-02-03 10:35:07-0500] [ALPM] upgraded elfutils (0.192-2 -> 0.192-3)
2025-02-03 10:35:07-0500] [ALPM] upgraded mpfr (4.2.1-4 -> 4.2.1-5)
2025-02-03 10:35:07-0500] [ALPM] upgraded gcc (14.2.1+r134+gab884fffe3fc-2 -> 14.2.1+r730+gc061ad5a36ba-1)
2025-02-03 10:35:07-0500] [ALPM] upgraded libsysprof-capture (47.2-1 -> 47.2-2)
2025-02-03 10:35:08-0500] [ALPM] upgraded libtool (2.5.4+r1+gbaa1fe41-1 -> 2.5.4+r1+gbaa1fe41-2)

--and--

2025-02-08 12:30:03-0500] [ALPM] upgraded glibc (2.41+r2+g0a7c7a3e283a-1 -> 2.41+r6+gcf88351b685d-1)
2025-02-08 12:30:04-0500] [ALPM] upgraded gcc-libs (14.2.1+r730+gc061ad5a36ba-1 -> 14.2.1+r753+g1cd744a6828f-1)
2025-02-08 12:30:04-0500] [ALPM] upgraded libelf (0.192-3 -> 0.192-4)
2025-02-08 12:30:04-0500] [ALPM] upgraded binutils (2.43_1+r186+g61f8adadd6db-1 -> 2.44-1)
2025-02-08 12:30:04-0500] [ALPM] upgraded elfutils (0.192-3 -> 0.192-4)
2025-02-08 12:30:04-0500] [ALPM] upgraded mpfr (4.2.1-5 -> 4.2.1-6)
2025-02-08 12:30:04-0500] [ALPM] upgraded gcc (14.2.1+r730+gc061ad5a36ba-1 -> 14.2.1+r753+g1cd744a6828f-1)
2025-02-08 12:30:04-0500] [ALPM] upgraded libsysprof-capture (47.2-2 -> 47.2-3)
2025-02-08 12:30:04-0500] [ALPM] upgraded libtool (2.5.4+r1+gbaa1fe41-2 -> 2.5.4+r1+gbaa1fe41-3)

EDIT: You could try downgrading this whole gang (those of which are relevant to your system) to the previous glibc 2.40.x.

I've not downgraded because a) the issues I'm seeing are intermittent, non-disruptive... and b) must be root-caused and solved moving forward.

I'll poke around upstreams.

Last edited by tekstryder (2025-02-15 05:09:01)

chrislg · 2025-02-15 08:59:21

Well I had issues with the current debian-stable and FreeBSD too. Debian freezes directly after sddm on the login screen. I think it is a hardware issue but I will try downgrading.

Last edited by chrislg (2025-02-15 10:54:48)

seth · 2025-02-15 09:07:01

seth wrote:

There's generally https://wiki.archlinux.org/title/Ryzen#Troubleshooting

chrislg · 2025-02-15 11:38:41

Enabled XMP-1 and PBO from BIOS, still no success. Will try the other options too and I will retest my memory with Memtest86+, this time with more passes...

Last edited by chrislg (2025-02-15 11:43:06)

seth · 2025-02-15 14:39:14

Enabled XMP-1 and PBO from BIOS

*en*abling that will hardly stabilize the system?

chrislg · 2025-02-15 16:17:27

If the memory voltage is lower wouldn't XMP help?

chrislg · 2025-02-15 20:53:00

https://imgur.com/a/f7tsrtA

loqs · 2025-02-15 22:18:55

What if disable PBO and any other overclocking settings while leaving the memory settings configured by XMP. If that has no effect try adjusting the memory settings to a slower JEDEC standard.

chrislg · 2025-02-15 23:27:56

Windows 11 Installation crashes with BSOD, that's ridiculous !!

chrislg · 2025-02-15 23:37:28

Here is the first kernel panic on January 29th

https://imgur.com/a/M3cpche

Some of the KWin crashes that I get eg from today:
https://imgur.com/a/VwBwmcq

Last edited by chrislg (2025-02-15 23:43:04)

seth · 2025-02-16 08:47:53

Windows 11 Installation crashes with BSOD, that's ridiculous !!

No, it's not.
Don't overclock the system.
Disable XMP, use teh most conservative RAM timings you can, disable PBO.
If that doesn't stabilize the system, start to feed more voltage to the CPU, but also just run memtes86+ for a day or so for good measure.

Standard disclaimer: 3rd link below. Mandatory.
Disable it (it's NOT the BIOS setting!) and reboot windows and linux twice for voodo reasons.

Arch Linux

#1 2025-02-08 14:49:41

System crashes directly during or after boot

#2 2025-02-09 17:18:16

Re: System crashes directly during or after boot

#3 2025-02-09 17:41:48

Re: System crashes directly during or after boot

#4 2025-02-09 18:43:44

Re: System crashes directly during or after boot

#5 2025-02-09 19:08:47

Re: System crashes directly during or after boot

#6 2025-02-11 14:39:21

Re: System crashes directly during or after boot

#7 2025-02-12 09:02:05

Re: System crashes directly during or after boot

#8 2025-02-12 09:43:08

Re: System crashes directly during or after boot

#9 2025-02-12 13:39:56

Re: System crashes directly during or after boot

#10 2025-02-12 20:22:00

Re: System crashes directly during or after boot

#11 2025-02-12 20:56:42

Re: System crashes directly during or after boot

#12 2025-02-12 22:14:15

Re: System crashes directly during or after boot

#13 2025-02-12 22:59:31

Re: System crashes directly during or after boot

#14 2025-02-14 23:39:58

Re: System crashes directly during or after boot

#15 2025-02-15 02:19:47

Re: System crashes directly during or after boot

#16 2025-02-15 08:59:21

Re: System crashes directly during or after boot

#17 2025-02-15 09:07:01

Re: System crashes directly during or after boot

#18 2025-02-15 11:38:41

Re: System crashes directly during or after boot

#19 2025-02-15 14:39:14

Re: System crashes directly during or after boot

#20 2025-02-15 16:17:27

Re: System crashes directly during or after boot

#21 2025-02-15 20:53:00

Re: System crashes directly during or after boot

#22 2025-02-15 22:18:55

Re: System crashes directly during or after boot

#23 2025-02-15 23:27:56

Re: System crashes directly during or after boot

#24 2025-02-15 23:37:28

Re: System crashes directly during or after boot

#25 2025-02-16 08:47:53

Re: System crashes directly during or after boot

Board footer