2

A I have to run the latest Kernel on my Acer Swift 3 (SF314-42-R2UX) to make suspend work I downloaded Kernel 5.15.13-051513 form mainline repository an signed it following the instructions here: https://ubuntu.com/blog/how-to-sign-things-for-secure-boot

And this:

https://gloveboxes.github.io/Ubuntu-for-Azure-Developers/docs/signing-kernel-for-secure-boot.html

I had to sign it as the Bios does not support to boot unsigned Kernel. I manged to sign the Kernels and for some day I could boot normally. But out of the sudden I cant boot anymore, I get:

ALERT! /dev/disk/by-uuid/xxxxxxxxx does not exist. Dropping to a shell
initramfs:_

I can boot to the original Ubuntu Kernel 5.13.0-24 though.

Can anybody give me a hint, what is going wrong here?

[UPDATE:] I found out, that this error is caused by running update-initramfs which is triggered by system updates. apt gives me this error:

update-initramfs: Generating /boot/initrd.img-5.15.13-051513-generic.signed
W: missing /lib/modules/5.15.13-051513-generic.signed
W: Ensure all necessary drivers are built into the linux image!
depmod: ERROR: could not open directory /lib/modules/5.15.13-051513-generic.signed: No such file or directory
depmod: FATAL: could not search modules: No such file or directory
cat: /var/tmp/mkinitramfs_3MmNVh/lib/modules/5.15.13-051513-generic.signed/modules.builtin: Datei oder Verzeichnis nicht gefunden
W: Can't find modules.builtin.modinfo (for locating built-in drivers' firmware, supported in Linux >=5.2)
depmod: WARNING: could not open modules.order at /var/tmp/mkinitramfs_3MmNVh/lib/modules/5.15.13-051513-generic.signed: No such file or directory
depmod: WARNING: could not open modules.builtin at /var/tmp/mkinitramfs_3MmNVh/lib/modules/5.15.13-051513-generic.signed: No such file or directory

And after that I can't boot anymore into the self signed kernel. Uninstalling and reinstalling and resigning it helps until next update.

What do I have to do, to avoid this problem?

[UPDATE2:] For self compiled kernels the problem seems to be solved by deactivating kernel modules: https://unix.stackexchange.com/a/685947/495409

user5950
  • 6,056
  • 11
  • 59
  • 83
  • 1
    What what?! I was using Debian 11 (default kernel 5.10), and I compiled kernel 5.15 after [downloading it](https://www.kernel.org/), and I got an almost identical error. Like your issue, my old 5.10 kernel works fine, but 5.15 doesn't work. I had the same issue, and [asked for help on Unix SE](https://unix.stackexchange.com/q/685843) (because I have Debian), but I can say that I have this issue too, so please @ ping me if you figure it out. Thanks! If it helps, I compiled mine with the default settings, and I didn't need to sign it. – cocomac Jan 11 '22 at 04:15
  • @cocomac Sure, I keep you up to date. It's a strange thing. What computer model are yu ousing? Did you have AMD related suspend issues like me (https://gitlab.freedesktop.org/drm/amd/-/issues/1721)? Or what made you compile a kernel? – user5950 Jan 11 '22 at 09:36
  • I didn't have a good reason for recompiling it... I had never (successfully) compiled a kernel before, and because mine way old, I though why not. So, I tried it. I'm using a Dell PowerEdge T610, with 2x Intel Xeon E5620s in it. It's a headless server, and I almost always manage it over SSH, though sometimes I do use a VGA display. – cocomac Jan 11 '22 at 17:24
  • Update: I figured it out. I had to disable kernel module support. See my full solution [here](https://unix.stackexchange.com/a/685947/495409). – cocomac Jan 11 '22 at 18:22

0 Answers0