Multiple application crashes with kernel 6.8.0-90

Recent updates included kernel 6.8.0-90, but after installing it and rebooting, a lot of things go awry:

  • Applications not starting (Slack)
  • Applications starting, but with partial functionality (Betterbird: cannot detect system theme anymore, hangs when opening link in email)
  • Applications crashing when performing certain actions (Terminator: crashes upon selecting text)
  • Gnome session crashing (not sure if it even logged me out, everything was gone quite suddenly and I was left with the gnome login screen).

This is accompanied by system journal entries like:

Dec 27 22:15:18 PC2021 kernel: [   47.708593] traps: goa-daemon[3590] trap invalid opcode ip:771064144a9b sp:7fff19acbb30 error:0 in libgio-2.0.so.0.7200.4[7710640a1000+112000]
Dec 27 22:15:18 PC2021 kernel: [   47.739631] traps: evolution-sourc[3562] trap invalid opcode ip:74472720ba9b sp:7ffcc6d9b190 error:0 in libgio-2.0.so.0.7200.4[744727168000+112000]
Dec 27 22:15:18 PC2021 kernel: [   47.750269] traps: goa-daemon[3604] trap invalid opcode ip:7b0f1c4cda9b sp:7ffe8e28e850 error:0 in libgio-2.0.so.0.7200.4[7b0f1c42a000+112000]
Dec 27 21:15:44 PC2021 kernel: [   73.721117] traps: gnome-software[3761] trap invalid opcode ip:70e4d22bda9b sp:7ffd0ddd6d40 error:0 in libgio-2.0.so.0.7200.4[70e4d221a000+112000]
Dec 27 21:17:19 PC2021 kernel: [  168.329427] traps: deja-dup-monito[4804] trap invalid opcode ip:7156638c1a9b sp:7ffe81ca9910 error:0 in libgio-2.0.so.0.7200.4[71566381e000+112000]
Dec 27 21:17:20 PC2021 kernel: [  169.753821] traps: xdg-desktop-por[4891] trap invalid opcode ip:76d362c4da9b sp:7fff587b03d0 error:0 in libgio-2.0.so.0.7200.4[76d362baa000+112000]
Dec 27 21:17:20 PC2021 systemd[3372]: xdg-desktop-portal.service: Main process exited, code=killed, status=4/ILL
Dec 27 21:17:20 PC2021 systemd[3372]: xdg-desktop-portal.service: Failed with result 'signal'.
Dec 27 21:17:20 PC2021 systemd[3372]: Failed to start Portal service.
Dec 27 21:22:54 PC2021 kernel: [  503.404866] traps: flatpak-portal[5032] trap invalid opcode ip:79f097419a9b sp:7ffee5a64340 error:0 in libgio-2.0.so.0.7200.4[79f097376000+112000]
Dec 27 21:22:54 PC2021 systemd[3372]: flatpak-portal.service: Main process exited, code=killed, status=4/ILL
Dec 27 21:22:54 PC2021 systemd[3372]: flatpak-portal.service: Failed with result 'signal'.
Dec 27 21:22:54 PC2021 systemd[3372]: Failed to start flatpak portal.
Dec 27 23:02:36 PC2021 kernel: [ 6485.626045] NVRM: VM: invalid mmap
(A whole lot of the above bunched together, followed by a couple of the error below)
Dec 27 23:02:36 PC2021 firefox.desktop[16505]: Exiting due to channel error.

This was terminator crashing:

Dec 27 23:13:03 PC2021 kernel: [ 7111.834587] terminator[4358]: segfault at 7ffd28210d80 ip 000079e3a2999282 sp 00007ffd03dc4680 error 6 in libgio-2.0.so.0.7200.4[79e3a28f7000+112000] likely on CPU 2 (core 2, socket 0)
Dec 27 23:13:03 PC2021 kernel: [ 7111.834613] Code: 53 48 83 ec 18 64 48 8b 04 25 28 00 00 00 48 89 44 24 08 31 c0 48 89 e3 e8 ab d0 ff ff 48 89 ef 48 89 c6 e8 00 40 f6 ff 48 9b <40> 20 8b 00 c7 44 24 04 04 00 00 00 89 04 24 eb 0f 0f 1f 44 00 00

and another:

Dec 27 23:13:39 PC2021 kernel: [ 7148.730995] terminator[18849]: segfault at 7ffc91708cb0 ip 00007fcd9902d282 sp 00007ffc6d2bc5b0 error 6 in libgio-2.0.so.0.7200.4[7fcd98f8b000+112000] likely on CPU 10 (core 2, socket 0)
Dec 27 23:13:39 PC2021 kernel: [ 7148.731021] Code: 53 48 83 ec 18 64 48 8b 04 25 28 00 00 00 48 89 44 24 08 31 c0 48 89 e3 e8 ab d0 ff ff 48 89 ef 48 89 c6 e8 00 40 f6 ff 48 9b <40> 20 8b 00 c7 44 24 04 04 00 00 00 89 04 24 eb 0f 0f 1f 44 00 00

Is something seriously broken in this version of the kernel?
At the moment, I've reverted to 6.8.0-88 and none of the above problems are present.

CPU: AMD Ryzen 7 3700X
Video: GeForce GTX 1060 6GB (using nvida-580 driver)
Mainboard: X570 Steel Legend (bios P5.60)
OS: Zorin 17.3 Core

Are you on XOrg or Wayland? Try to switch to Xorg. Logout and at login click on your username then a cog wheal appears at the bottom right corner. There select "Zorin Desktop on Xorg"

1 Like

I'm on Wayland. At some point (possibly the 17.2 to 17.3 upgrade), zorin switched to wayland and there is no Xorg option anywhere.

Do you use a login without password? If yes, there is no cogwheel shown. You can edit the file

/etc/gdm3/custom.conf

and remove the # at line
# WaylandEnable=false
and then run

sudo systemctl restart gdm3
1 Like

Doing that results in an empty textmode screen with a cursor topleft (and that is with the "good working" -88 kernel. Running that with the -90 kernel will likely not improve that.

I would add the Xorg log, but I can't seem to attach textfiles, and putting everything under a collapsable section makes the forum complain about having more than 5 links in it (?).

It seems my Xorg config doesn't know about the 2 screens I have attached. No idea how I can fix that (or how I broke that in the first place).

With some AI assistance, I managed to get that working again. I can confirm that the errors with regard to applications and gnome itself are not showing themselves (yet).

However, I am seeing this in my dmesg:

Dec 29 22:17:45 PC2021 systemd-udevd[642]: Worker [667] terminated by signal 11 (SEGV)
Dec 29 22:17:45 PC2021 kernel: systemd-udevd[667]: segfault at 73cc722012d3 ip 000073cc689f82f0 sp 00007ffd657541b0 error 4 in libkmod.so.2.3.7[73cc689f2000+11000] likely on CPU 0 (core 0, socket 0)
Dec 29 22:17:45 PC2021 kernel: Code: 04 25 28 00 00 00 0f 85 1e 01 00 00 48 81 c4 38 02 00 00 4c 89 e0 5b 5d 41 5c 41 5d 41 5e 41 5f c3 66 0f 1f 84 00 00 00 00 00 <0f> be 13 0f be 43 01 4c 8d 6b 02 89 c6 29 d0 89 d1 8d 50 01 4c 63
Dec 29 22:17:45 PC2021 systemd-udevd[642]: LNXSYSTM:00: Worker [667] failed
Dec 29 22:17:45 PC2021 systemd-udevd[642]: Worker [675] terminated by signal 11 (SEGV)
Dec 29 22:17:45 PC2021 systemd-udevd[642]: LNXPWRBN:00: Worker [675] failed
Dec 29 22:17:45 PC2021 kernel: systemd-udevd[645]: segfault at 73cc722012d3 ip 000073cc689f82f0 sp 00007ffd657541d0 error 4 in libkmod.so.2.3.7[73cc689f2000+11000] likely on CPU 5 (core 5, socket 0)
Dec 29 22:17:45 PC2021 kernel: Code: 04 25 28 00 00 00 0f 85 1e 01 00 00 48 81 c4 38 02 00 00 4c 89 e0 5b 5d 41 5c 41 5d 41 5e 41 5f c3 66 0f 1f 84 00 00 00 00 00 <0f> be 13 0f be 43 01 4c 8d 6b 02 89 c6 29 d0 89 d1 8d 50 01 4c 63
Dec 29 22:17:45 PC2021 kernel: systemd-udevd[675]: segfault at 73cc722012d3 ip 000073cc689f82f0 sp 00007ffd657541b0 error 4 in libkmod.so.2.3.7[73cc689f2000+11000] likely on CPU 13 (core 5, socket 0)
Dec 29 22:17:45 PC2021 kernel: Code: 04 25 28 00 00 00 0f 85 1e 01 00 00 48 81 c4 38 02 00 00 4c 89 e0 5b 5d 41 5c 41 5d 41 5e 41 5f c3 66 0f 1f 84 00 00 00 00 00 <0f> be 13 0f be 43 01 4c 8d 6b 02 89 c6 29 d0 89 d1 8d 50 01 4c 63
Dec 29 22:17:45 PC2021 kernel: systemd-udevd[647]: segfault at 73cc722012d3 ip 000073cc689f82f0 sp 00007ffd657541d0 error 4 in libkmod.so.2.3.7[73cc689f2000+11000] likely on CPU 10 (core 2, socket 0)
Dec 29 22:17:45 PC2021 kernel: Code: 04 25 28 00 00 00 0f 85 1e 01 00 00 48 81 c4 38 02 00 00 4c 89 e0 5b 5d 41 5c 41 5d 41 5e 41 5f c3 66 0f 1f 84 00 00 00 00 00 <0f> be 13 0f be 43 01 4c 8d 6b 02 89 c6 29 d0 89 d1 8d 50 01 4c 63
Dec 29 22:17:45 PC2021 kernel: systemd-udevd[651]: segfault at 73cc722012d3 ip 000073cc689f82f0 sp 00007ffd657541d0 error 4 in libkmod.so.2.3.7[73cc689f2000+11000] likely on CPU 2 (core 2, socket 0)
Dec 29 22:17:45 PC2021 kernel: Code: 04 25 28 00 00 00 0f 85 1e 01 00 00 48 81 c4 38 02 00 00 4c 89 e0 5b 5d 41 5c 41 5d 41 5e 41 5f c3 66 0f 1f 84 00 00 00 00 00 <0f> be 13 0f be 43 01 4c 8d 6b 02 89 c6 29 d0 89 d1 8d 50 01 4c 63
Dec 29 22:17:45 PC2021 kernel: systemd-udevd[650]: segfault at 73cc722012d3 ip 000073cc689f82f0 sp 00007ffd657541d0 error 4 in libkmod.so.2.3.7[73cc689f2000+11000] likely on CPU 3 (core 3, socket 0)
Dec 29 22:17:45 PC2021 kernel: Code: 04 25 28 00 00 00 0f 85 1e 01 00 00 48 81 c4 38 02 00 00 4c 89 e0 5b 5d 41 5c 41 5d 41 5e 41 5f c3 66 0f 1f 84 00 00 00 00 00 <0f> be 13 0f be 43 01 4c 8d 6b 02 89 c6 29 d0 89 d1 8d 50 01 4c 63
Dec 29 22:17:45 PC2021 kernel: systemd-udevd[681]: segfault at 73cc722012d3 ip 000073cc689f82f0 sp 00007ffd657541d0 error 4 in libkmod.so.2.3.7[73cc689f2000+11000] likely on CPU 4 (core 4, socket 0)
Dec 29 22:17:45 PC2021 kernel: systemd-udevd[687]: segfault at 73cc722012d3 ip 000073cc689f82f0 sp 00007ffd657541d0 error 4
Dec 29 22:17:45 PC2021 kernel: Code: 04 25 28 00 00 00 0f 85 1e 01 00 00 48 81 c4 38 02 00 00 4c 89 e0 5b 5d 41 5c 41 5d 41 5e 41 5f c3 66 0f 1f 84 00 00 00 00 00 <0f> be 13 0f be 43 01 4c 8d 6b 02 89 c6 29 d0 89 d1 8d 50 01 4c 63
Dec 29 22:17:45 PC2021 kernel:  in libkmod.so.2.3.7[73cc689f2000+11000] likely on CPU 0 (core 0, socket 0)

That doesn't seem too healthy either. No segfaults are logged at all when using 6.8.0-88.

Segmentation faults are usually indications of programs that have been badly written and trying to access memory that they shouldn't, and rarely, potential hardware failure. I would run Disks and use the S.M.A.R.T. test on your hard drive.

I know. The thing is that these segfaults do not happen at all when using kernel 6.8.0-88, but only started happening when the upgrade to 6.8.0-90 was done. And they do not happen when rebooting to 6.8.0-88 again.

I think this highlights another issue. It is amazing how kernels have a limited life from a security angle when the Linux kernel seems more concerned with keeping up with new hardware. It feels to me like the Linux kernel is trying to keep up with that other OS to placate gamers at the expense of a useable system.

Update: I left my system running on 6.8.0-90 since the last post. When returning to it, I found myself unable to login, the mouse was responding to movement and clicks (though that was harder to verify). Since I locked the system when leaving it yesterday, I had to login, however the login prompt behaved like the [Enter]-key was stuck: it continuously said authentication failed (erratically blinking that message). Keypressed did not result in blobs appearing in the password box. CIFS shares ceased working at some point, SSH didn't work either. The system was still pingable.

I was forced to do a hard-reset, which ended me up in the recovery console upon the next boot, since there were filesystem errors in the root filesystem (ext4) that needed manual oversight. Not that I could do much at that point but tell fsck to fix everything anyway.
It seems like I've lost the entire system journal since my last post and God knows what else.

What strikes me as odd is that I would have expected more reports about this to pop up on the internet, but we'll see.

If someone knows anything I could do to test to narrow down what exactly is going wrong, please do speak up :slight_smile:

All I can say is that whilst other users may not have had segfaults, there has been instances where a newer kernel has meant the system does not work properly - this is why it is good to have the GRUB menu shown (I had to use Boot Repair with Zorin 18 to get the GRUB menu to show!). Once you have GRUB menu present at boot you can select 'Advanced options' and then choose an earlier kernel. The next step is to prevent the system from installing the same problematical kernel in future updates.

1 Like