Boot freezes at splash screen, no ttys - failing on 3 kernels (6.5, 6.2, 5.25) & recovery mode boots

Help
,
1

I’m having an issue booting. For kernel version 6.5, 6.2, and 5.25 that I have available, I can’t boot normally nor in recovery mode. In each case, boot freezes at the loading “splash screen”. The KDE gear freezes, I cannot yet access other ttys/ptys, and boot proceeds no further.

I’m seeing kernel UBSAN errors, it looks like these are unrelated.

I’ve seen others report issues with the lockscreen version.

They mention kdecoration, and I’ve tried downgrading the related lockscreen package for neon/debian packaged kde.

This has not worked for me.

1 Like
2

I’ve let the KDE Partition Manager run a e2fsck -f -y -v /dev/mapper/luks-... of the disk, it didn’t detect any issues, one of the first things that comes to mind. I suppose more subtle SSD corruption is still possible, just affecting a package. Seems likely to be fixable with reinstalling packages which I’ll be attempting anyway—proceeding with the assumption of some package version or hardware conflict.

I’m booting the live image on the USB I installed from in 2022, to diagnose issues.

I’m curious about ways to debug the bringup of services, increase verbosity of logging at boot. I have journals from the failing boots, but nothing stands out in journals, searching /fail|error (ignoring case) nor reading through it.

I’m running a MacBookPro11,5 (2015 retina). There are wake from suspend issues, there are wifi limitations.

kdecoration mentioned in the arch forum.

libkdecorations2-5v5/jammy,now 4:5.27.10-0xneon+22.04+jammy+release+build28 amd64 [installed,automatic]
libkdecorations2private10/jammy,now 4:5.27.10-0xneon+22.04+jammy+release+build28 amd64 [installed,automatic]
libkscreenlocker5/jammy,now 5.24.4-0ubuntu1 amd64 [installed,upgradable to: 5.27.10-0xneon+22.04+jammy+release+build33]

I downgraded libkscreenlocker5 and I should have tried downgrading libkdecorations2-5v5, trying now.

3

At the moment, I think my best route forward involves the minimal way to downgrade libkdecorations2-5v5 to version 5.24.4-0ubuntu1. However, its dependency libkdecorations2private9, seems to

version conflict: `$ sudo apt install libkdecorations2-5v5=4:5.24.4-0ubuntu1` 
The following packages have unmet dependencies:
 libkdecorations2-5v5 : Depends: libkdecorations2private9 (= 4:5.24.4-0ubuntu1) but 4:5.27.10-0xneon+22.04+jammy+release+build28 is to be installed
 neon-desktop : Depends: breeze but it is not going to be installed
                Depends: kinfocenter but it is not going to be installed
                Depends: kwin-wayland but it is not going to be installed
                Depends: kwin-x11 but it is not going to be installed
                Depends: sddm-theme-breeze but it is not going to be installed
                Recommends: kde-config-gtk-style but it is not going to be installed
                Recommends: kwin-addons but it is not going to be installed
                Recommends: neon-apport but it is not going to be installed
                Recommends: plasma-dataengines-addons but it is not going to be installed
                Recommends: plasma-wallpapers-addons but it is not going to be installed
                Recommends: plasma-widgets-addons but it is not going to be installed
                Recommends: plasma-workspace-wayland but it is not going to be installed
 plasma-desktop : Depends: breeze (>= 4:5.14) but it is not going to be installed
                  Depends: plasma-integration (>= 5.14) but it is not going to be installed
                  Depends: plasma-workspace (>= 4:5.14) but it is not going to be installed
                  Recommends: kde-config-gtk-style (>= 4:5.14) but it is not going to be installed
                  Recommends: kinfocenter (>= 4:5.14) but it is not going to be installed
                  Recommends: kwin-x11 (>= 4:5.14) but it is not going to be installed or
                              kwin
E: Error, pkgProblemResolver::Resolve generated breaks, this may be caused by held packages.

Still have conflicts with sudo apt install libkdecorations2{-5v5,private9}=4:5.24.4-0ubuntu1.

It looks like neon-desktop is the root here, plasma-desktop is automatic from that. So I’ll pick an earlier version of neon-desktop.

root@neon:/var/debug-issues# apt list -a neon-desktop
Listing... Done
neon-desktop/jammy,now 4+p22.04+vrelease+git20240125.1419 amd64 [installed]
neon-desktop/jammy 4+p22.04+vrelease+git20240122.0627 amd64
neon-desktop/jammy 4+p22.04+vrelease+git20240119.2255 amd64
neon-desktop/jammy 4+p22.04+trelease+git20230711.0933 amd64
4

No improvement. I’ve installed previous neon-desktop versions, and rebooted. Possibly critically, though apropos of nothing for the moment, I (re)installed the plasma-dataengines-addons, plasma-wallpapers-addons, plasma-widgets-addons. I remember needing to install them for a splash screen I used at one point. I guess that nice splash screen will be available to me to configure again, once I can again finish booting my installed system.

5

Maybe not strange at all, since all of those are discontinued.
5.15, 6.1 and 6.6 would be a better fit, since they are the long term ones.

Please see this for reference.
https://kernel.org/

Disclaimer.
I do not know how neon handles kernels and kde.
But it would make sense that the absolute latest software would also require latest or LTS kernels.

Yeah, no, don’t do that (look at arch wiki).
You are on a debian based system. (I assumed kde neon, ubuntu fork IIRC, but maybe I misinterpreted that). Or wait, that’s kubuntu I am thinking of.
Kinda hard to answer without info.

1 Like
6

I appreciate the engagement, effort! Parsing your response in order.

Ok, supporting your hint about kernel EOL support. KDE Plasma version support lists 5.27 as the current version, 5.27.10 latest as of 05 Dec 2023, and 5.24.7 is their previous LTS. (My live disk appears to be 5.26.2, not supported, but also not relevant.)

Re: arch forums, I searched far and wide for “splash screen” issues, as KDE is careful to call it, and found a simliar sounding issue with a recent arch version of a very plausible packaged shared on my system. Arch and debian versions seem to have slightly different naming conventions, but it’s the same components.

Perhaps I don’t need to be on neon. What sources/packaging do you build your system from, @bedna?

LSB & Kernel versions

For futher debugging, I’m very open to suggestions about where else I can check versions, hopefully rolling back or otherwise try to debug. I do wish I had boot-into-vm ready, it would be nice to not have to re-initialize the chroot and re-login to this discuss forum after each live boot.

Sorry that my debugging and reporting so far is not as linear as it could be.

I’m not sure what other information might be helpful.
In debug chroot:

  • lsb
$ lsb_release -a
Distributor ID: Neon
Description:    KDE neon 5.27
Release:        22.04
Codename:       jammy
  • Installed kernel versions, exact kernel-version strings,
# strings /boot/vmlinuz* | grep -e -generic 
5.15.0-92-generic (buildd@lcy02-amd64-005) #102-Ubuntu SMP Wed Jan 10 09:33:48 UTC 2024
5.15.0-92-generic (buildd@lcy02-amd64-005) (gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0, GNU ld (GNU Binutils for Ubuntu) 2.38) #102-Ubuntu SMP Wed Jan 10 09:33:48 UTC 2024
5.15.0-92-generic (buildd@lcy02-amd64-005) #102-Ubuntu SMP Wed Jan 10 09:33:48 UTC 2024
5.15.0-92-generic (buildd@lcy02-amd64-005) (gcc (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0, GNU ld (GNU Binutils for Ubuntu) 2.38) #102-Ubuntu SMP Wed Jan 10 09:33:48 UTC 2024
6.2.0-36-generic (buildd@lcy02-amd64-050) #37~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Mon Oct  9 15:34:04 UTC 2
6.2.0-36-generic (buildd@lcy02-amd64-050) (x86_64-linux-gnu-gcc-11 (Ubuntu 11.4.0-1ubuntu1~22.04) 11.4.0, GNU ld (GNU Binutils for Ubuntu) 2.38) #37~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Mon Oct  9 15:34:04 UTC 2
6.5.0-15-generic (buildd@bos03-amd64-040) #15~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Fri Jan 12 18:54:30 UTC 2
6.5.0-15-generic (buildd@bos03-amd64-040) (x86_64-linux-gnu-gcc-12 (Ubuntu 12.3.0-1ubuntu1~22.04) 12.3.0, GNU ld (GNU Binutils for Ubuntu) 2.38) #15~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Fri Jan 12 18:54:30 UTC 2
6.5.0-15-generic (buildd@bos03-amd64-040) #15~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Fri Jan 12 18:54:30 UTC 2
6.5.0-15-generic (buildd@bos03-amd64-040) (x86_64-linux-gnu-gcc-12 (Ubuntu 12.3.0-1ubuntu1~22.04) 12.3.0, GNU ld (GNU Binutils for Ubuntu) 2.38) #15~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Fri Jan 12 18:54:30 UTC 2

Why kernel, package versions are available / where from

Interesting, those aren’t what’s available for me, running (thinking aloud for you, me, others—and for corrigibility)

  • KDE’s neon, which pushes more fresh KDE WM/DM/app packages but inherits from (and has infrequent snapshots pulled into…)
  • ubuntu, which pushes some many of its own kernel and core lib packages but inherits a lot from
  • debian (sid?) or other release repos.
Filtering through `apt list linux-* --installed` 
linux-base/jammy,now 4.5ubuntu9 all [installed,automatic]
linux-firmware/now 20220329.git681281e4-0ubuntu3.26 all [installed,local]
linux-generic-hwe-22.04/now 6.5.0.15.15~22.04.8 amd64 [installed,local]
linux-generic/now 5.15.0.92.89 amd64 [installed,local]
linux-headers-5.15.0-92-generic/now 5.15.0-92.102 amd64 [installed,local]
linux-headers-5.15.0-92/now 5.15.0-92.102 all [installed,local]
linux-headers-5.19.0-50-generic/now 5.19.0-50.50 amd64 [installed,local]
linux-headers-6.2.0-36-generic/now 6.2.0-36.37~22.04.1 amd64 [installed,local]
linux-headers-6.5.0-15-generic/now 6.5.0-15.15~22.04.1 amd64 [installed,local]
linux-headers-generic-hwe-22.04/now 6.5.0.15.15~22.04.8 amd64 [installed,local]
linux-headers-generic/now 5.15.0.92.89 amd64 [installed,local]
linux-hwe-5.19-headers-5.19.0-50/now 5.19.0-50.50 all [installed,local]
linux-hwe-6.2-headers-6.2.0-36/now 6.2.0-36.37~22.04.1 all [installed,local]
linux-hwe-6.5-headers-6.5.0-15/now 6.5.0-15.15~22.04.1 all [installed,local]
linux-image-5.15.0-92-generic/now 5.15.0-92.102 amd64 [installed,local]
linux-image-6.2.0-36-generic/now 6.2.0-36.37~22.04.1 amd64 [installed,local]
linux-image-6.5.0-15-generic/now 6.5.0-15.15~22.04.1 amd64 [installed,local]
linux-image-generic-hwe-22.04/now 6.5.0.15.15~22.04.8 amd64 [installed,local]
linux-image-generic/now 5.15.0.92.89 amd64 [installed,local]
linux-libc-dev/now 5.15.0-92.102 amd64 [installed,local]
linux-modules-5.15.0-92-generic/now 5.15.0-92.102 amd64 [installed,local]
linux-modules-6.2.0-36-generic/now 6.2.0-36.37~22.04.1 amd64 [installed,local]
linux-modules-6.5.0-15-generic/now 6.5.0-15.15~22.04.1 amd64 [installed,local]
linux-modules-extra-5.15.0-92-generic/now 5.15.0-92.102 amd64 [installed,local]
linux-modules-extra-6.2.0-36-generic/now 6.2.0-36.37~22.04.1 amd64 [installed,local]
linux-modules-extra-6.5.0-15-generic/now 6.5.0-15.15~22.04.1 amd64 [installed,local]
linux-sound-base/jammy,now 1.0.25+dfsg-0ubuntu7 all [installed,automatic]
a selection of latest I have available from `# apt list linux-headers-*` 
linux-headers-5.19.0-50-generic/now 5.19.0-50.50 amd64 [installed,local]
linux-headers-6.0.0-1006-oem/jammy-updates 6.0.0-1006.6 amd64
linux-headers-6.0.0-1007-oem/jammy-updates,jammy-security 6.0.0-1007.7 amd64
linux-headers-6.0.0-1008-oem/jammy-updates,jammy-security 6.0.0-1008.8 amd64
linux-headers-6.0.0-1009-oem/jammy-updates 6.0.0-1009.9 amd64
linux-headers-6.0.0-1010-oem/jammy-updates,jammy-security 6.0.0-1010.10 amd64
linux-headers-6.1.0-1004-oem/jammy-updates,jammy-security 6.1.0-1004.4 amd64
linux-headers-6.2.0-36-generic/now 6.2.0-36.37~22.04.1 amd64 [installed,local]
linux-headers-6.5.0-15-generic/now 6.5.0-15.15~22.04.1 amd64 [installed,local]

It looks like ubuntu, or maybe kde, don’t care to release on top of LTS kernel versions, except perhaps for their own LTS releases. Ubuntu offers xx.04 versions, every other year, as their LTS release: (20.04 focal*,) 22.04 jammy, (24.04 noble* coming).

Takeaway so far

It seems likely this is not a kernel issue, the previous kernels I have installed, I previously booted from. What’s different now, is the KDE desktop packages. I have mention in passing of a splash screen / lock screen issue from the arch forums. I’ve tried re-installing the packages, but it’s nothing so simple.

It’s frustrating, I have boot logs, there’s little remarkable in the journals nor at the end of the journals, kernel modules being loaded, at worst, maybe
systemd-modules-load fails occasionally on ddcci, only into the 6.5 kernel, but it’s loaded in other kernels, so not the issue here.

My experience trying to log in entails: reaching the splash screen, having the gear animation stop, and hearing the fans continue to rev up while other ttys are inaccessible and nothing happens on the splash screen.

7

I am confused. You cited kernels in your initial post, and now only versions of kde. Kde is not a kernel.

Ubuntu.

I can tell you this much. Ubuntu has software in their repositories that are years old, sometimes even drivers.

To solve this issue, you should probably restore to whatever ubuntu distributes in their repositories, and stay on that. Because I suspect without you providing this information, that you have installed something outside their repos.

If you want to run the latest versions of anything, do not use Ubuntu.
I would even go so far as to argue not to use a debian based distro for anything related to “latest versions”.

In the future If you are unsure, a simple inxi or even neofetch are great tools for providing basic info in a nice format for you to post.

Or maybe I am missunderstanding everything and all you have done is a sudo apt update && sudo apt upgrade and suddenly could not boot.
If that is the case, a Ubuntu forum is most likely way more fitting for this issue, since it is “their” package you install.

Edit
I did a tiny deeper look into what KDE neon actually is, and found this at the bottom of their faq about updating.

If you prefer to use the command line you can use the pkcon command:

  • pkcon refresh
  • pkcon update

This will install all new packages and uses the same PackageKit code as Plasma Discover. Some uses of apt do not install new packages which makes it less suitable for KDE neon.

8

I appreciate the continued digging. Unfortunately, pkcon also doesn’t want to operate in a chroot environment.

Following up on this thread, I’ve followed the advice of others.

In case updates were interrupted, dpkg --configure -a then apt install to fix.

update-initramfs -u -k [version] then update-grub.

https://www.reddit.com/r/kde/comments/13plvkk/my_computer_was_updating_with_terminal_but_i_had/

On using this forum & package freshness

It feels appropriate to post here about packages closely related to the default installed packages, and for standard additional configuring of the flagship base system. It’s a supported configuration, encouraged, all-but-endorsed in the neon faq you found. I didn’t share your appreciation for the breadth of this community, many ways people get their software to use KDE.

It appears sddm packaged version, 0.20.0, is as current as the most recent git release Releases · sddm/sddm · GitHub.

Status update

I notice my 6.5 kernel boots are reporting issues loading ddcci and facetimehd modules. The boot logs for 6.2 and 5.25 kernels stop within a line or two after this, but

A friend suggested I check the apt history log,

/var/log/apt/history.log 
Start-Date: 2024-02-14  13:47:03
Commandline: /usr/bin/apt install libpcap-dev libev-dev libnl-3-dev libnl-genl-3-dev libnl-route-3-dev
Requested-By: xenon (1000)
Install: libnl-3-dev:amd64 (3.5.0-0.1), libnl-genl-3-dev:amd64 (3.5.0-0.1), libnl-route-3-dev:amd64 (3.5.0-0.1), libev4:amd64 (1:4.33-1, automatic), libev-d
ev:amd64 (1:4.33-1)
End-Date: 2024-02-14  13:47:06

Start-Date: 2024-02-17  13:29:41
Commandline: /usr/bin/apt install apport
Requested-By: xenon (1000)
Install: apport-symptoms:amd64 (0.24, automatic), apport:amd64 (2.20.11-0ubuntu82.3), python3-systemd:amd64 (234-3ubuntu2, automatic)
Remove: neon-apport:amd64 (0.0+p22.04+trelease+git20220901.0903)
End-Date: 2024-02-17  13:29:46

After this I just bounce between apport and (null) neon-apport, learning, before attempts to install specific versions of libkscreenlocker5 vs kde-config-screenlocker, and neon-desktop , which due to debian’s package versioning management has little to no influence on which version of the screenlocker is installed, any of them satisfy, so I need to pin specific version of the package I want to check for issues, or maybe a direct r/dependency.

6.5 kernel boot, tail of log -- UBSAN radoen driver errors 
Feb 18 15:11:25 kernel: microcode: updated early: 0x19 -> 0x1c, date = 2019-11-12
Feb 18 15:11:25 kernel: Linux version 6.5.0-15-generic (buildd@bos03-amd64-040) (x86_64-linux-gnu-gcc-12 (Ubuntu 12.3.0-1ubuntu1~22.04) 12.3.0, GNU ld (GNU Binutils for Ubuntu) 2.38) #15~22.04.1-Ubuntu SMP PREEMPT_DYNAMIC Fri Jan 12 18:54:30 UTC 2 (Ubuntu 6.5.0-15.15~22.04.1-generic 6.5.3)
Feb 18 15:11:25 kernel: Command line: BOOT_IMAGE=/boot/vmlinuz-6.5.0-15-generic root=UUID=f73f2537-1bd4-4991-81ee-843dcad71a9e ro loglevel=7 luks.crypttab=no
...
Feb 18 15:11:25 kernel: ACPI: BIOS _OSI(Darwin) query honored via DMI
Feb 18 15:11:25 kernel: ACPI: [Firmware Bug]: BIOS _OSI(Linux) query ignored
Feb 18 15:11:25 kernel: ACPI Error: Needed type [Reference], found [Integer] (____ptrval____) (20230331/exresop-66)
Feb 18 15:11:25 kernel: ACPI Error: AE_AML_OPERAND_TYPE, While resolving operands for [Store] (20230331/dswexec-433)
Feb 18 15:11:25 kernel: 
Feb 18 15:11:25 kernel: No Local Variables are initialized for Method [_PDC]
Feb 18 15:11:25 kernel: 
Feb 18 15:11:25 kernel: Initialized Arguments for Method [_PDC]:  (1 arguments defined for method invocation)
Feb 18 15:11:25 kernel:   Arg0:   (____ptrval____) <Obj>           Buffer(12) 01 00 00 00 01 00 00 00
Feb 18 15:11:25 kernel: 
Feb 18 15:11:25 kernel: ACPI Error: Aborting method \_PR.CPU0._PDC due to previous error (AE_AML_OPERAND_TYPE) (20230331/psparse-529)
...
Feb 18 15:11:25 kernel: ================================================================================
Feb 18 15:11:25 kernel: UBSAN: array-index-out-of-bounds in /build/linux-hwe-6.5-BV4m5T/linux-hwe-6.5-6.5.0/drivers/gpu/drm/radeon/radeon_atombios.c:2717:34
Feb 18 15:11:25 kernel: index 48 is out of range for type 'UCHAR [1]'
Feb 18 15:11:25 kernel: CPU: 6 PID: 160 Comm: systemd-udevd Not tainted 6.5.0-15-generic #15~22.04.1-Ubuntu
Feb 18 15:11:25 kernel: Hardware name: Apple Inc. MacBookPro11,5/Mac-06F11F11946D27C5, BIOS MBP114.88Z.0184.B00.1806051659 06/05/2018
Feb 18 15:11:25 kernel: Call Trace:
Feb 18 15:11:25 kernel:  <TASK>
Feb 18 15:11:25 kernel:  dump_stack_lvl+0x48/0x70
Feb 18 15:11:25 kernel:  dump_stack+0x10/0x20
Feb 18 15:11:25 kernel:  __ubsan_handle_out_of_bounds+0xc6/0x110
Feb 18 15:11:25 kernel:  radeon_atombios_parse_power_table_6+0x39c/0x3d0 [radeon]
Feb 18 15:11:25 kernel:  radeon_atombios_get_power_modes+0x1e9/0x210 [radeon]

I remember booting successfully despite seeing these errors. And I saw some people recompile kernels, with minimal patches, a few lines of array[1] → array[].

6.5 kernel boot, tail of log -- nothing after this, no errors -- most logs look like this 
Feb 18 15:01:27 kernel: EXT4-fs (dm-0): re-mounted f73f2537-1bd4-4991-81ee-843dcad71a9e r/w. Quota mode: none.
Feb 18 15:01:27 systemd[1]: Mounted Kernel Debug File System.
Feb 18 15:01:27 kernel: lp: driver loaded but no devices found
Feb 18 15:01:27 systemd[1]: Mounted Kernel Trace File System.
Feb 18 15:01:27 systemd[1]: Finished Create List of Static Device Nodes.
Feb 18 15:01:27 kernel: ppdev: user-space parallel port driver
Feb 18 15:01:27 systemd[1]: modprobe@configfs.service: Deactivated successfully.
Feb 18 15:01:27 systemd[1]: Finished Load Kernel Module configfs.
Feb 18 15:01:27 systemd[1]: modprobe@drm.service: Deactivated successfully.
Feb 18 15:01:27 systemd[1]: Finished Load Kernel Module drm.
Feb 18 15:01:27 systemd[1]: modprobe@fuse.service: Deactivated successfully.
Feb 18 15:01:27 systemd[1]: Finished Load Kernel Module fuse.
Feb 18 15:01:27 systemd[1]: Finished Remount Root and Kernel File Systems.
Feb 18 15:01:27 systemd[1]: Finished Set the console keyboard layout.
Feb 18 15:01:27 systemd-journald[398]: Journal started
Feb 18 15:01:27 systemd-journald[398]: Runtime Journal (/run/log/journal/5b8b96b856894331a42209fb115daca5) is 8.0M, max 158.6M, 150.6M free.
Feb 18 15:01:27 systemd-modules-load[421]: Inserted module 'lp'
Feb 18 15:01:27 kernel: mc: Linux media interface: v0.10
Feb 18 15:01:27 kernel: pstore: Registered efi_pstore as persistent store backend
Feb 18 15:01:27 kernel: videodev: Linux video capture interface: v2.00
Feb 18 15:01:27 systemd-modules-load[421]: Inserted module 'ppdev'
Feb 18 15:01:27 systemd-modules-load[421]: Inserted module 'parport_pc'
Feb 18 15:01:27 systemd-modules-load[421]: Module 'i2c_dev' is built in
Feb 18 15:01:27 systemd-modules-load[421]: Inserted module 'msr'
Feb 18 15:01:27 systemd-modules-load[421]: Failed to find module 'ddcci'
Feb 18 15:01:27 systemd[1]: Mounted NFSD configuration filesystem.
Feb 18 15:01:27 systemd[1]: Started Journal Service.
Feb 18 15:01:27 kernel: facetimehd: loading out-of-tree module taints kernel.
Feb 18 15:01:27 kernel: facetimehd: module verification failed: signature and/or required key missing - tainting kernel
Feb 18 15:01:27 kernel: facetimehd 0000:05:00.0: Found FaceTime HD camera with device id: 1570
Feb 18 15:01:27 kernel: facetimehd 0000:05:00.0: Setting 64bit DMA mask
Feb 18 15:01:27 systemd[1]: Finished Monitoring of LVM2 mirrors, snapshots etc. using dmeventd or progress polling.
Feb 18 15:01:27 systemd[1]: modprobe@efi_pstore.service: Deactivated successfully.
Feb 18 15:01:27 systemd[1]: Finished Load Kernel Module efi_pstore.
Feb 18 15:01:27 systemd[1]: Mounting FUSE Control File System...
Feb 18 15:01:27 systemd[1]: Mounting Kernel Configuration File System...
Feb 18 15:01:27 kernel: facetimehd 0000:05:00.0: S2 PCIe link init succeeded
Feb 18 15:01:27 kernel: facetimehd 0000:05:00.0: Refclk: 25MHz (0xa)
Feb 18 15:01:27 systemd[1]: Starting Flush Journal to Persistent Storage...
Feb 18 15:01:27 systemd[1]: Condition check resulted in Platform Persistent Storage Archival being skipped.
Feb 18 15:01:27 systemd[1]: Starting Load/Save Random Seed...
Feb 18 15:01:27 systemd-journald[398]: Time spent on flushing to /var/log/journal/5b8b96b856894331a42209fb115daca5 is 8.670ms for 1501 entries.
9

It seems like I should request that kde-neon be released on top of LTS kernel releases.

Someone reporting the same issue I have with the AMD driver on 1/30.

https://bugzilla.kernel.org/show_bug.cgi?id=218435

KDE Neon has released on

  • a 5.15.0 kernel, good, last LTS
  • Then a 6.2 kernel, not good, a minor version after an LTS kernel, 6.1, so support is no longer available.
  • 6.5, not good, a minor version before 6.6 LTS kernel. 6.5 is already out of support.

Underlying releases for reference. I’m not sure of KDE developer policy on LTS versus temporarily up to date releases:

10

You should not be getting UBSan errors at all. From what you have said I suspect a corrupted filesystem on your disk. There can be several source of this that I know of. In case of the iommu unmapping ssd pages before they’re written, you can boot with iommu=pt and reinstall the system. You could also have a bad RAM overclock.

11

They are, on the faq:

It uses the foundation of the latest Ubuntu LTS (22.04).

They use UBUNTU releases. If you are frustrated with their release schedule and what kernels they use, you should take it up with Ubuntu.
22.04 is apparently the latest LTS available according to Releases - Ubuntu Wiki

1 Like
12

Correct me if I’m wrong, but the UBSAN errors appear related to ongoing intentional changes in the kernel, switching to use of dynamic arrays. IIRC, I saw these errors on successful boots before my issue(s) started.

(details … expandable section)

Would I expect fsck to report no errors? I’m using ext4 (in luks), so no validation, could be file data block errors. Wouldn’t I expect to be able to fix this issue by recreating the initramfs, update-initramfs -c -k all?

After updating initramfs and update grub boot options, including Debugging options, I notice regular boots do not begin syystemd service initialization.

[ (boot ts)] async_fx: api initialized (async)

https://i.fluffy.cc/9xGThvTppnCKzpMgXKF46LTNCv7F07XP.html

Trying

amdgpu.dc=0
13

UBSAN can also be caused by hardware errors… hardware behaving in unusual ways.

But you don’t need metadata corruption in your filesystem for it to be corrupt. When I had problems with one of my machines, there was a bad filesystem write while updating the gcc binary – I could not perform LTO or compile a kernel without gcc crashing. The fault was not with gcc, I needed a complete reinstall with some kernel options active to fix it.

If you had a bad write while updating the kernel or some other vital system component such as systemd, you could get random freezes during the init process or the ubsan errors.

Because you are getting different behaviour after doing exactly the same thing (duplicating the last initram update), would seem to indicate something is wrong with your disk or memory.

1 Like