Gnome crashes

Ask Fedora
1

Hi there,

Lately I was experiencing many freezes/crashes on my computer.
It appears that changing a bios option (related to the disk - Intel VMD) after I had Fedora installed was the cause..
So I just went ahead and reinstalled Fedora from scratch.
At first I was so happy as it seemed I had no crash.

But after using for a bit, I started getting crashes again :frowning:
It makes my computer unsable as it appears realtively quickly.

Usually my screen freezes and I can’t do anything (like switching to TTY) but hard reboot the computer. Sometimes my mouse cursor is moving (still no TTY).

But in the last crash something weird happened. I was able to TTY and kill gnome-shell then somehow I was able to log back in but the only thing I was seeing is an error message with a sad emoji icon (fullscreen with white background) asking to log out and log back in.
But I was still seeing the same message until it froze for sure and had to hard reboot.

Something worth noting is that I installed X11 and ran under X11 as the error happens even quicker on Wayland.

Here are the journalctl for 4 sessions. I’m joining only with errors to make it more readable but also the warnings in case it’s useful.

The last one under X11 which gave me access to TTY:

Wayland with complete freeze (no mouse no TTY) - Here the error one is not showing anything useful (AFAICT):

X11 with complete freeze (no TTY):

X11 with complete freeze (no TTY):

I already did test the memory and the SSD which don’t have errors. It’s a brand new computer and Windows is working just well.

I can give more details if necessary also I don’t care with my current install so I can do many tests if necessary.

Thanks for any help.

PS: Right now (when writing this post) it appears it’s still not crashing.. And I only have a terminal and brave browser running.
On the previous sessions I was always running an electron app - I’m pretty sure this is related. How can I be sure and instead go in the Electron repo and report there?

2

My computer just crashed again when I was just installing inxi.. I figured out that on many posts inxi output is posted..
So it crashed when using the terminal.
The error output journalctl -b -1 -p3 is not showing anything at the time it happened: UNTITLED - Pastebin Service

So here is the warning output -p4 (which, oddly enough, has refereneces to xwayland although I’m running in X11): UNTITLED - Pastebin Service

EDIT:
BTW, here is my inxi -Fzxx:

System:
  Kernel: 6.15.8-200.fc42.x86_64 arch: x86_64 bits: 64 compiler: gcc v: 15.1.1
  Desktop: GNOME v: 48.3 tk: GTK v: 3.24.49 wm: gnome-shell dm: GDM
    Distro: Fedora Linux 42 (Workstation Edition)
Machine:
  Type: Detachable System: ASUSTeK product: ASUS Zenbook Duo UX8406CA_UX8406CA
    v: 1.0 serial: <superuser required>
  Mobo: ASUSTeK model: UX8406CA v: 1.0 serial: <superuser required>
    UEFI: American Megatrends LLC. v: UX8406CA.308 date: 05/16/2025
Battery:
  ID-1: BAT0 charge: 54.4 Wh (73.2%) condition: 74.3/75.1 Wh (99.0%)
    volts: 17.0 min: 16.0 model: ASUSTeK ASUS Battery serial: N/A
    status: charging
CPU:
  Info: 16-core model: Intel Core Ultra 9 285H bits: 64 type: MCP
    arch: Arrow Lake rev: 2 cache: L1: 1.6 MiB L2: 28 MiB L3: 24 MiB
  Speed (MHz): avg: 400 min/max: 400/5400:4500:2500 cores: 1: 400 2: 400
    3: 400 4: 400 5: 400 6: 400 7: 400 8: 400 9: 400 10: 400 11: 400 12: 400
    13: 400 14: 400 15: 400 16: 400 bogomips: 117964
  Flags: avx avx2 ht lm nx pae sse sse2 sse3 sse4_1 sse4_2 ssse3 vmx
Graphics:
  Device-1: Intel Arrow Lake-P [Intel Graphics] vendor: ASUSTeK driver: i915
    v: kernel arch: Xe-LPG ports: active: eDP-1 off: eDP-2
    empty: DP-1,DP-2,HDMI-A-1 bus-ID: 00:02.0 chip-ID: 8086:7d51
  Device-2: Shinetech USB2.0 FHD UVC WebCam driver: uvcvideo type: USB
    rev: 2.0 speed: 480 Mb/s lanes: 1 bus-ID: 3-9:4 chip-ID: 3277:0055
  Display: x11 server: X.Org v: 21.1.18 with: Xwayland v: 24.1.8
    compositor: gnome-shell driver: X: loaded: modesetting alternate: fbdev,vesa
    dri: iris gpu: i915 display-ID: :0 screens: 1
  Screen-1: 0 s-res: 1920x1200 s-dpi: 96
  Monitor-1: eDP-1 pos: primary model: Samsung 0x419d res: N/A dpi: 242
    diag: 356mm (14")
  Monitor-2: eDP-2 note: disabled model: Samsung 0x419d res: N/A dpi: 242
    diag: 356mm (14")
  API: OpenGL v: 4.6 vendor: intel mesa v: 25.1.4 glx-v: 1.4 es-v: 3.2
    direct-render: yes renderer: Mesa Intel Graphics (ARL) device-ID: 8086:7d51
  API: EGL Message: EGL data requires eglinfo. Check --recommends.
  Info: Tools: api: glxinfo x11: xdriinfo, xdpyinfo, xprop, xrandr
Audio:
  Device-1: Intel vendor: ASUSTeK driver: sof-audio-pci-intel-mtl
    bus-ID: 00:1f.3 chip-ID: 8086:7728
  API: ALSA v: k6.15.8-200.fc42.x86_64 status: kernel-api
  Server-1: PipeWire v: 1.4.7 status: active with: 1: pipewire-pulse
    status: active 2: wireplumber status: active 3: pipewire-alsa type: plugin
    4: pw-jack type: plugin
Network:
  Device-1: Intel driver: iwlwifi v: kernel port: N/A bus-ID: 00:14.3
    chip-ID: 8086:7740
  IF: wlo1 state: up mac: <filter>
  IF-ID-1: br-f8199b6127cf state: down mac: <filter>
  IF-ID-2: docker0 state: down mac: <filter>
Bluetooth:
  Device-1: Intel driver: btusb v: 0.8 type: USB rev: 2.0 speed: 12 Mb/s
    lanes: 1 bus-ID: 3-10:5 chip-ID: 8087:0037
  Report: btmgmt ID: hci0 rfk-id: 3 state: up address: <filter> bt-v: 5.4
    lmp-v: 13
Drives:
  Local Storage: total: 953.87 GiB used: 60.74 GiB (6.4%)
  ID-1: /dev/nvme0n1 vendor: Western Digital model: WD PC SN5000S
    SDEQNSJ-1T00-1002 size: 953.87 GiB speed: 63.2 Gb/s lanes: 4
    serial: <filter> temp: 35.9 C
Partition:
  ID-1: / size: 551.3 GiB used: 60.36 GiB (10.9%) fs: btrfs dev: /dev/dm-0
    mapped: luks-fa45b65f-4bba-4938-81b6-d321c94e0dd6
  ID-2: /boot size: 920.7 MiB used: 340.8 MiB (37.0%) fs: ext4
    dev: /dev/nvme0n1p5
  ID-3: /boot/efi size: 996 MiB used: 50.3 MiB (5.1%) fs: vfat
    dev: /dev/nvme0n1p1
  ID-4: /home size: 551.3 GiB used: 60.36 GiB (10.9%) fs: btrfs
    dev: /dev/dm-0 mapped: luks-fa45b65f-4bba-4938-81b6-d321c94e0dd6
Swap:
  ID-1: swap-1 type: zram size: 8 GiB used: 0 KiB (0.0%) priority: 100
    dev: /dev/zram0
Sensors:
  System Temperatures: cpu: N/A mobo: N/A
  Fan Speeds (rpm): cpu: 2743
Info:
  Memory: total: 32 GiB note: est. available: 30.71 GiB used: 3.98 GiB (13.0%)
  Processes: 475 Power: uptime: 6m wakeups: 0 Init: systemd v: 257
    target: graphical (5) default: graphical
  Packages: pm: rpm pkgs: N/A note: see --rpm pm: flatpak pkgs: 8
    Compilers: N/A Shell: fish v: 4.0.2 running-in: kitty inxi: 3.3.38
3

For clarification.
This appears to be one of the devices that can be used in laptop mode or separately in tablet mode. Is that correct?

The asus site shows that model as dual screen and detachable keyboard.
It seems the hardware design and the fact that it has only been certified on windows 11 (with a comment that it may not work on windows 10) are very new, so suitable drivers may not be available.

Asus also shows the release date for that model as February 2025.

4

Well.. I’m not sure if tablet is the correct word, but it has 2 built-in monitors indeed (quite fancy TBH).
The bottom one can be covered by the keyboard.
When the keyboard is attached it acts like a regular laptop.
When the keyboard is detached I can enable the second screen and have like a dual monitor setup. In this case the keyboard should be connected in bluetooth. But I’m just using a regular USB keyboard and don’t bother with connecting the keyboard (at least on Linux). On windows it seemless but I don’t bother enabling-disabling manually on Linux (I have some scripts where I run gdctl accordingly)

EDIT: Just saw your update. Yes it’s fairly new, but I don’t care not having special features working (like the keyboard etc.). There are already some people using it with success e.g. GitHub - ywzjackal/zenbook-duo-2025-linux: zenbook-duo-2025-linux utils but I just one the basic working.
On my previous install (kernel version 6.15.5) I was seeing journalctl errors related to i915 but here I don’t see them anymore which is a good sign I want to say..

5 
1_p3:Aug 03 14:24:49 kernel: i915 0000:00:02.0: [drm] *ERROR* GT0: rcs0 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
1_p3:Aug 03 14:24:49 kernel: i915 0000:00:02.0: [drm] *ERROR* GT0: rcs0 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
1_p4:Aug 03 14:24:49 kernel: i915 0000:00:02.0: [drm] *ERROR* GT0: rcs0 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
1_p4:Aug 03 14:24:49 kernel: i915 0000:00:02.0: [drm] *ERROR* GT0: rcs0 reset request timed out: {request: 00000001, RESET_CTL: 00000001}
1_p4:Aug 03 14:30:56 kernel: i915 0000:00:02.0: [drm] PHY A failed to request refclk after 1us.
6

The 6.15.8 kernel xe module (modinfo xe| grep 7D51) seems to be appropriate, but there may be an issue (thermal management?) that precludes using it.

Since you have new hardware, you should participate in F43 testing to make sure your graphics will work in F43.

7

Hi..

I did notice many errors (and not always the same) related to i915.

I also tried the xe module instead of the i915 (using force_probe in GRUB)

But unfortunately, my crashes is still happening..

Is there some kind of setting I can set in GRUB related to GUC, or GT0 or GT1?

8

Hi,

I already tested xe instead of i915 (using the GRUB options which I confirmed using lspci -k showing driver in use: xe.

I guess I can give it a try to F43..

Something interesting is that it happens that my unit also crashes randomly on Windows…

All BSOD..

So I wonder if there’s a hardware issue? The RAM and the SSD is all good however. Both the diagnostic tool of the BIOS and memtest/smartctl on Linux succeeded.

I’m very sad as I can’t use it at all.. After a few minutes this happens.. Quicker when using Electron apps.. But this might be just a coincidence..

9

There will be a kernel test week soon for the 6.16 also on F42

p.s.
do not forget to recheck the https://linux-hardware.org/ site to see if there are issues you especially have to look for.

1 Like
10

Thanks for the probe link!

Here is my probe: HW probe of ASUSTek ASUS Zenbook Duo UX8... #6e05e7805d

It doesn’t seem to have any issues?

In order to limit the causes…

I formatted my whole disk and get rid of windows and just installed Fedora alone. I also reverted the VMD option to be enabled (as it was originally).

I still get instant freezes upon the first boot.

But I can definitely see some kind of graphic artifacts.

Something odd is that it doesn’t happen at all in the live session..

I found some i915 options for GRUB as well as some nvme_core but I’m not sure what I should use exactly..
I tried some random ones, but either it’s the same or worse..

11

You could investigate booting into a live session and see what drivers you are using there, first a sudo dnf search intel- also to see the packages installed.. Then see what you get when booting without a live session. inxi -Fzxx is a command which helps you to see this in a overview. The boot options you can check with grubby

All three commands should give enough output to find the packages/drivers you will need to install after a new installation with a live iso. As an alternative you can also test the network-installation which is a bit slower but should download more you also have on a live session.

Also important is to check if you have the newest bios installed.
fwupdmgr get-devices to see if there is something new.

12

You can compare the output from infix -Fzxx from the Live USB session (you will have to sudo dnf install inxi, and note that the package won’t be saved across a reboot) with the output from the installed system.

I’m a bit concerned that kernels before 6.16 may be missing power management changes related to xe graphics. Depending on your environment, you should consider running a temperature monitoring app.

13

Thank you for providing detailed information. A few comments:

dmesg.1 shows errors while handling the PCI bus for VMD. They may be insignificant. To reduce complexity l would disable VMD again.

[    1.883981] pci 10000:e0:06.2: bridge window [io  size 0x1000]: can't assign; no space
[    1.884055] pci 10000:e0:06.2: bridge window [io  size 0x1000]: failed to assign

dmesg.1 shows an unhandled page fault caused by the i915 driver:

[  110.253311] BUG: unable to handle page fault for address: ffffb4beb7d21000
[  110.253336] #PF: supervisor write access in kernel mode
[  110.253480] #PF: error_code(0x0002) - not-present page

dmesg.1 The name of a kernel parameter is wrong. The correct name is nvme_core.default_ps_max_latency_us:

[    1.884961] nvme_core: unknown parameter 'default_ps_max_us' ignored

smartctl shows the ssd drive has reached the critical temp of 88 c once:

Warning  Comp. Temperature Time:    1
Critical Comp. Temperature Time:    1
1 Like
14

The bug looks similar to Making sure you're not a bot!, fixed in Jan 2025. I guess the fix should be in kernel 6.15. The hw probe is with kernel 6.14.0.

1 Like