Good to know some people have success. 
2 Likes
Switched to sddm but alas this also doesn’t work with WM and 5.12.x!
Strange! I haven’t had any issue on my two systems. One is Intel with Nvidia on Xfce. The other is Ryzen with amdgpu and i have Kde, Xfce and Cinnamon desktops all separate triple boot and have no issue.
It seems only to affect Window Managers. Desktop Environments are fine.
Yes… i just checked both Cinnamon and Xfce and i didn’t have them set to true but changed it anyway. I definitely had it set to true on the Nvidia machine with Xfce.
Any commonality of compositor with those WM’s? As opposed to the functioning DE’s? Can a DE compositor be swapped in on any of them for testing/elimination?
Of coursze - I know next to nothing about WM’s - too old to learn them now that they are good enough to use 
A coredumping Xorg could be due to a partial upgrade.
Alternatively, something like https://bugs.archlinux.org/task/70726 could be related, therefore the workaround is to use linux-lts until the regression is fixed.
1 Like
Fairly sure picom is the only compositor I have installed.
Some DEs come with their compositor built-in to their WM - like XFCE4-wm. Can it be swapped in to the other WM’s as a compositor only? Not sure how they are specced in their setup…
The Arch reddit is full of this very issue today.
1 Like
Also worth trying in this situation:
sudo chown -R $USER $HOME
sudo rm $HOME/.Xauthority
(Second one only applies if you’re using Xorg, and obviously try one at a time to check which one had an effect.).
1 Like
Ouch. Same problem but deeper than others - BSPWM edition here.
I can only login to tty. Any commands are not accessible. So can’t do cat, ls, cd, cp, mv, vi, vim, uname, rm, sudo etc. Kernel 5.12.2-arch1-1.
Should I already install a fresh system?
I would follow the wiki here (EnOS) on arch-chroot from a live boot - then downgrade the kernel (probably to 5.11.16). Instructions for that are in the wiki as well, I think - but it is easy - just pick the local kernel that is the closest match, and say yes to ‘Ignore’.
From there (after a reboot) you should be running, and can investigate if something else is wrong.
1 Like
This is why you keep two kernels installed - whatever you fancy and linux-lts.
Given how many people don’t do this I kind of wonder whether it should be a standard inclusion in the installer…
9 Likes
Maybe because you have to learn some grub to get the default boot to be the one you mostly want? With rEFInd I just change the order of listed kernels in refind.conf…
1 Like
Ah, yes indeedy. If you have linux-lts and linux installed then GRUB will default to linux-lts. 
Worth reading if anyone is in that position: https://wiki.archlinux.org/title/GRUB/Tips_and_tricks#Multiple_entries
Yup - that’s why I abandoned -lts back then. Now I have -zen, -lts, and linux all awaiting if needed (and the settings are accessible from any build OR live because it’s on /boot/efi/EFI No need to chroot to make to make a change.
1 Like
I used this tutorial when I installed lts kernel so that I could switch between it and others I had installed.
EDIT: Remember to install the appropriate linux headers though. I forgot and it’s the reason I broke VMWare, once I figured that out all is well and I can easily switch for testing this current issue.
Another alternative is to use the AKM (kernel manager) GUI provided with EnOS. Choosing what to boot depends on your boot method, of course…
I switched to linux-lts kernel and now logs me in. The zen kernel doesn’t work and might try the harden kernel since I’m trying to get more into security side of IT. Will the mainstream kernel be fixed for this issue or is it EOS only issue?
edit: linux-lts isn’t detecting my wifi card
ifconfig -a
no interface