Reloading system manager configuration... system freeze, after reboot os is broken

This is the second time that my whole system hard freezes on pacman -Syu and I have to restart by holding down the power button (SysRQ was not working, so the kernel was well and truly frozen)

It was on “Reloading system manager configuration…” when it happened.

But after I rebooted my system is now completely broken, with error messages along the lines of ‘failed to contact dbus daemon’ and ‘failed to connect to the system bus, connection refused’ if i try to run a command after logging in it says something like connection refused as well. It’s so bad that to shutdown the PC i have to use sysrq because it just gives me an error if I try to shutdown the normal way.

What do I do? What caused this? How can it be prevented? Do I really have to reinstall the whole thing to fix this? Is my drive corrupted or something? (It seemed to pass fsck…)

Journalctl log at time of crash

Mar 09 17:35:11 Murmaider sudo[1063179]:   rabcor : TTY=pts/0 ; PWD=/home/rabcor ; USER=root ; COMMAND=/usr/bin/pacman -Syyu
Mar 09 17:35:11 Murmaider sudo[1063179]: pam_unix(sudo:session): session opened for user root(uid=0) by rabcor(uid=1000)
Mar 09 17:35:44 Murmaider systemd-logind[1118]: /efi/loader/loader.conf:5: Unknown line 'reboot-for-bitlocker', ignoring.
Mar 09 17:35:44 Murmaider rtkit-daemon[1313]: Successfully made thread 1114778 of process 1105454 owned by '1000' high priority at nice level -10.
Mar 09 17:35:44 Murmaider rtkit-daemon[1313]: Supervising 8 threads of 5 processes of 1 users.
Mar 09 17:35:44 Murmaider rtkit-daemon[1313]: Successfully made thread 1114779 of process 1105454 owned by '1000' high priority at nice level -10.
Mar 09 17:35:44 Murmaider rtkit-daemon[1313]: Supervising 9 threads of 5 processes of 1 users.
Mar 09 17:35:44 Murmaider rtkit-daemon[1313]: Successfully made thread 1114831 of process 1105454 owned by '1000' high priority at nice level -10.
Mar 09 17:35:44 Murmaider rtkit-daemon[1313]: Supervising 10 threads of 5 processes of 1 users.
Mar 09 17:35:53 Murmaider pressure-vessel-wrap[1125518]: W: Failed to load Vulkan ICD #1 from /usr/share/vulkan/icd.d/radeon_icd.i686.json: openat(/usr/share/vulkan/icd.d/radeon_icd.i686.json): No such file or directory
Mar 09 17:36:05 Murmaider pressure-vessel-wrap[1135155]: W: Failed to load Vulkan ICD #1 from /usr/share/vulkan/icd.d/radeon_icd.i686.json: openat(/usr/share/vulkan/icd.d/radeon_icd.i686.json): No such file or directory
Mar 09 17:36:11 Murmaider pressure-vessel-wrap[1140132]: W: Failed to load Vulkan ICD #1 from /usr/share/vulkan/icd.d/radeon_icd.i686.json: openat(/usr/share/vulkan/icd.d/radeon_icd.i686.json): No such file or directory
Mar 09 17:36:23 Murmaider pressure-vessel-wrap[1150177]: W: Failed to load Vulkan ICD #1 from /usr/share/vulkan/icd.d/radeon_icd.i686.json: openat(/usr/share/vulkan/icd.d/radeon_icd.i686.json): No such file or directory
Mar 09 17:38:34 Murmaider dbus-broker-launch[1089]: Noticed file-system modification, trigger reload.
Mar 09 17:38:34 Murmaider dbus-broker-launch[1089]: Looking up NSS group entry for 'sudo'...
Mar 09 17:38:34 Murmaider dbus-broker-launch[1089]: NSS returned no entry for 'sudo'
Mar 09 17:38:34 Murmaider dbus-broker-launch[1089]: Invalid group-name in /usr/share/dbus-1/system.d/asusd.conf +9: group="sudo"
Mar 09 17:38:34 Murmaider dbus-broker-launch[1089]: Invalid group-name in /usr/share/dbus-1/system.d/org.supergfxctl.Daemon.conf +9: group="sudo"
Mar 09 17:38:34 Murmaider dbus-broker-launch[1089]: Noticed file-system modification, trigger reload.
Mar 09 17:38:34 Murmaider dbus-broker-launch[1089]: Looking up NSS group entry for 'sudo'...
Mar 09 17:38:34 Murmaider dbus-broker-launch[1089]: NSS returned no entry for 'sudo'
Mar 09 17:38:34 Murmaider dbus-broker-launch[1089]: Invalid group-name in /usr/share/dbus-1/system.d/asusd.conf +9: group="sudo"
Mar 09 17:38:34 Murmaider dbus-broker-launch[1089]: Invalid group-name in /usr/share/dbus-1/system.d/org.supergfxctl.Daemon.conf +9: group="sudo"

Update: Replacing dbus-broker-units with dbus-daemon-units allowed my system to boot properly (networking works, commands that didn’t before now work again, and i don’t get any errors until after I actually finish booting), however it broke a lot of packages so the system is still more or less unusable, and if i try to reinstall the libraries the packages are complaining about i just get ‘already exists in filesystem’ errors from pacman and it refuses to do anything (I tried to just delete the files to reinstall them, but then the broken packages just give other errors anyways)

Reinstalling dbus-broker-units puts me back to square one, so i’m pretty sure it’s configuration got corrupted somehow.

Use your live usb, chroot into your system and run the updates from there.

Hopefully, it will fix the issue.

It won’t hurt to run a fsck on your system partition(s) in case the hard shutdown has caused some filesystem corruption.

It’s not working.

The error at the end is normal in chroot. It has no relevance.

From the tiny part of the terminal output you have posted, it seems that the update has gone through and initramfs has been created.

So I would exit chroot and reboot.

I tried from chroot running pacman -Syu, reinstalling the kernels too, but there was no change :frowning:

1 Like

Is there any way I can see exactly what caused the crash at least, so if i reinstall i can prevent it from happening again? I mean “Reloading system manager configuration” does something specific, and there might be logs or something i can read right?

I’m afraid I don’t know what exactly may have caused this freeze on “Reloading system manager configuration” step so I don’t really know where to look for some info.

Here is a link you can have a look at for looking at some logs:
https://discovery.endeavouros.com/forum-log-tool-options/how-to-include-systemlogs-in-your-post/2021/03/

Stand by for other users to come by and hopefully provides more insight on this specific issue.

The pacman logs weren’t very helpful, they contained what looks like incoherent screaming at the time of the crash.

Well the only logs that seemed useful were from journalctl, so i added them to op…

Seems whatever caused the crash, dbus-broker-launch was responsible for it.

Ok i’m getting somewhere, i replaced dbus-broker-units with dbus-daemon-units, and the OS booted normally, however many packages (including my compositor) are broken and throwing errors about not being able to load libraries that are actually installed.

Btw, are you using EXT4 or Btrfs?
Just wondering since chroot instructions are slightly different for each one.

Also, here is a link to what I think could help in this case:

https://wiki.archlinux.org/title/Pacman#Pacman_crashes_during_an_upgrade

ext4

I would suggest a chroot and reinstall of all your native packages in case the pacman crash and the following hard shutdown may have caused some corruption.

Chroot and do pacman -Qqn | pacman -S -

https://wiki.archlinux.org/title/Pacman/Tips_and_tricks#Reinstalling_all_packages

doesn’t work…

...exists in filesystem
Errors occurred, no packages were upgraded.

I get like 5 million ‘exists in filesystem’ errors .

Then the “nuclear” option is to overwrite them all!

It comes with a warning as shown in the link I posted in my previous post.

However there doesn’t seem to be other options at this juncture, so as “the last resort” , in chroot:

pacman -S $(pacman -Qqn) --overwrite '*'

ok this worked, i’m back in my system and everything seems to be working, i’m using dbus-daemon-units though, i think i’ll just stick with that honestly.

So to make the entire solution in one post:

dbus-broker was somehow causing the issue for me.

Chrooting in and replacing dbus-broker-units package with dbus-daemon-units allowed me to boot properly but broke a couple packages.

Chrooting in and running pacman -S --overwrite "*" $(pacman -Qqn) allowed me to reinstall all packages and now my system works again (but many system config files got overwritten)

I still get issues though, with pacman.

(3/3) Checking which packages need to be rebuilt
pactree: error while loading shared libraries: libalpm.so.14: cannot open shared object file: No such file or directory
pactree: error while loading shared libraries: libalpm.so.14: cannot open shared object file: No such file or directory
pactree: error while loading shared libraries: libalpm.so.14: cannot open shared object file: No such file or directory
pactree: error while loading shared libraries: libalpm.so.14: cannot open shared object file: No such file or directory
pactree: error while loading shared libraries: libalpm.so.14: cannot open shared object file: No such file or directory
pactree: error while loading shared libraries: libalpm.so.14: cannot open shared object file: No such file or directory
pactree: error while loading shared libraries: libalpm.so.14: cannot open shared object file: No such file or directory
pactree: error while loading shared libraries: libalpm.so.14: cannot open shared object file: No such file or directory
pactree: error while loading shared libraries: libalpm.so.14: cannot open shared object file: No such file or directory
pactree: error while loading shared libraries: libalpm.so.14: cannot open shared object file: No such file or directory

Indeed i only have libalpm.so.13

Follow https://wiki.archlinux.org/title/Pacman#Pacman_crashes_during_an_upgrade

At step 4: pacman --sysroot /mnt -S pacman as root.

Doing that didn’t actually work, it just complained about the mirrors, but something i did must have worked because it is now fixed, honestly it might have been that i chrooted in and did a pacman -Syu which updated pacman-contrib? well idk, anyhow it is fine now.