So let me get this straight: with
I only achieve that there is no error message, but still the last selected kernel is not booted. How also, if that is avoided with it. I have the last LTS kernel as fallback 8at 1st and 2nd place in the grubloader, at 3rd place the CURRENT kernel to be booted. If I
GRUB_DEFAULT=2, the error message comes again. Sh. btrfs …
The reason is the grub.env which is on a filesystem not writable by grub and grub config defaults to writing the selected entry to grub.env. To work around it edit your /etc/default/grub and change from
OK, I’ve been there, so there’s no point in changing the grub.config. But how can I achieve that the kernel I preset boots?
As soon as I write something other than 0 on GRUB_DEFAULT, I get the error message back. I’ve been searching the net for a solution for several hours now. Or I kick btrfs in the a… and take ext4 again.Too bad, I wanted it for the snapshots …
I probably already did everything wrong during the setup. I should have created subvolumes, but my swap partition is still ext4. I’ll go to sleep now and think about it tomorrow, if the whole effort is worth it.
Well - immediately systemd boot and rEFInd come to mind (rEFInd for me). However, if you use grub to select between Timeshift snapshots, you might want to keep it. Of course, you can use rEFInd to boot the grub - but I’m not sure it gets around the difficulty you are having.
The reason for the message originates in the fact that grub cannot write to grub environment file (/boot/grub/grubenv) - I am not using grub - so I may remember the name incorrect - but grub writes the last used entry to the environment file and because grub cannot write on btrfs you get this grub message.
Only way around it is to have your /boot mounted on a ext filesystem or disable the save function.