This morning I reported a bug to the Debian maintainers for the problems encountered after the "systemd" update to version "255~rc2" (https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1056034).

They replied that it depends on the kernel I use (Liquorix) and in fact I have verified that with the Debian kernel, now at version 6.5, the problems do not occur. I checked and that's it.

Can it be solved?

Thank you,
Antonio
Back to top

The issue is being tracked here: github.com/damentz/liquorix-package/issues/147

Unfortunately, I believe this is a bug in systemd, but the systemd developers believe it's a bug in Liquorix.
Back to top

I have tried all bootable kernels, from version "6.4.15-2" to version "6.5.11-4" on debian/stable installation, but I always get the same errors.

This makes me suppose that it is no error in the Liquorix kernel and more generally to the latest changes made to the main stream kernel.

Then I tried to compile kernels "6.6" and "6.5.11" from "kernel.org" and they work fine: this, however, seems to indicate a problem in some patch applied to the Liquorix kernel...
Back to top

Can you try the latest kernel pushed up for Liquorix on v6.6.1? In my own testing this issue appears to be resolved.
Back to top

I updated the kernel to version "6.6.1-1-liquorix-amd64", no errors detected, even systemd services now work without errors.

The only warnings I see with "dmesg" are these:
----
[Sun Nov 19 09:47:22 2023] hid_bpf: Error preloading dispatcher HID BPF: -22
[Sun Nov 19 09:47:22 2023] RAS: Correctable Errors collector initialized.
[Sun Nov 19 09:47:22 2023] x86/split lock detection: #AC: Kernel crash on kernel split_locks and warning on userspace split_locks
[Sun Nov 19 09:47:22 2023] Specter V2: WARNING: Unprivileged eBPF is enabled with eIBRS active, data leaks via Specter v2 BHB attacks are possible!
[Sun Nov 19 09:47:25 2023] ACPI Warning: \_SB.PC00.XHCI.RHUB.HS14._DSM: Argument # type mismatch. 4 - Found [integer], ACPI requires [package] (20230628/nsarguments-61)
----
but they were there before.

Technical curiosity, what was the problem?
Back to top

I'm really not sure, there must have been a "fix" or stable patch that broke support for this next version of systemd. That or the systemd developers genuinely wrote some code that _will_ break on some kernels, we just won't see the damage until some time later.

Either way, marking this as resolved.
Back to top