After updating today, I rebooted. Since then, there is a black screen after the bios splash screen. I don’t reach the image selection prompt that would usually come after that. When I try entering tty mode with ctrl+alt+f1, nothing happens(I tried with other f keys as well, same result). I can boot onto the live environment on the ISO.
I checked a few logs(/var/log/pacman.log, /var/log/Xorg.0.log and journalctl) but saw mothing unusual, though I don’t really know anything about troubleshooting those applications beyond checking if anything in the log says error.
4 machines this morning went in the same situation after the daily rotation and update. Those 4 had newer BIOS available, so I tried to update them. It worked on 2 of them. (i3 12100f intel and B660 asrock motherboards with recent basic & economical AMD gpus).
The other 2 ( i3 10100f intel ) with older GPUs (zotak Nividia 610 passive) never came up to life and stayed in the same as you are describing. The indicator was just before the OS came to life, there was a message that said these old 610 geforce were not UEFI compatible anymore …
I decided to shuffle around 2 more recent amd gpus from 2 other machines who have onboard Cpu displays (ryzen 7600x and intel i5 11600).
I was able to go in the BIOSs of those two, but not the OS. So I reinstalled with the new Galileo release (which is possible to avoid if you chroot your system and repair it … )
More knowledgeable folks will come in to help you. That’s the most I can do in information transfer.
Since you can’t get into the system, you could mention, as best you can, all the hardware that is in your machine.
The knowledgeable members that will want to push you further in the repair will need the maximum you can provide.
I will stay logged, to learn the most about your repair …
edit: maybe the “nouveau” driver temporally - to get out the Nvidia problems that a lot of members seems to have. ( I only use nouveau on 9 machines )
From my understanding, if it was the nvidia driver, you would get far enough to boot normally, and enter tty, you just would not necessarily be able to get actual graphics.
Anyway, chrooting into my system and reverting everything all packages I have to yesterday’s version did not do anything.
Edit: I’m giving up for today, I’m going to bed. Maybe I’ll get to play the new path of exile league tomorrow
the first step is to get system back to boot properly so you can go to revert to nouveau driver to see if that makes you able to get X working again or at least TTY login.
You can use nvidia-inst tool to do so from chroot: nvidia-inst -n
Same also in case of login to TTY?
Would be helpful to get boot journal: journalctl -b -0 | eos-sendlog
but do not post the URL here better send me in a P.M. so you do not post details public.
Signed up to add this is happening to me too. Just installed endeavouros a few days ago. Everything was working well. Systemd updated and I had a notification to reboot which I delayed. Just rebooted now and total black screen after bios. Cannot access tty, nothing.
I can boot into another arch installation on a different disk.
Setup is an i5-12600kf, RTX 3080, Nvidia 545.29.06, kernel 6.6.4
Oh, right, that’s probably because EOS uses its own systemd-boot+dracut configuration now, so I guess its building Unified Kernel Images in some different way.
Have you tried to regenerate the UKIs with sudo reinstall-kernels?
But anyway, I think the problem is the same as in the linked issue, systemd update messed up creating the UKIs. (initrds don’t get included in the image)
I would try to downgrade systemd instead of reinstalling.
I am also having this issue after updating today on my other machine. I arch-chroot’ed into the system and tried the following:
switch to nouveau drivers with nvidia-inst -n; this made no difference.
downgrade systemd to 254, and running reinstall-kernels afterwards; this did not help either.
updated systemd back to 255, added uki_generator=ukify to /etc/kernel/install.conf, then reinstall-kernels; didn’t help either. I tried the same but setting the flag to dracut instead, which also did nothing.
I’m not very knowledgeable about dracut and creating boot images. Is there anything else I can try? Thanks.
Hey, I am not sure there is anything more you can try, maybe ping the devs for more help. I am not using EOS anymore (on “pure” Arch currently), so I am not sure what the situation is with systemd-boot + dracut, or if EOS is using UKIs by default at all, so the bug linked might not be relevant.
What you can try to do is: 1. check if your /efi partition is mounted correctly before running reinstall-kernels 2. check if you can boot with fallback menu entry 3. check if your initrds/UKIS are generated correctly (check their sizes, and use lsinitrd or objdump to inspect the contents of initrd/UKI for your kernels on efi partition).
Thanks for the help. My efi partition was mounted correctly afaik, and I can’t boot with the fallback entry either. I had a look with lsinitrd but don’t really know what I’m looking for, though the sizes seemed plausible to me.
Not sure what to try next as it seems my timeshift never created any backups, even though I thought I scheduled it a few months ago. I wouldn’t mind reinstalling but if anyone has more ideas I would gladly hear them.
