I’m curious to know why F42 uses a separate /boot partition even when FDE isn’t used.
I’ve read the news about F43 using a 2go /boot partition by default, I’ve read users complain that the their /boot is full, so why use a separate partition when grub2 can read /boot in a btrfs partition?
I know that openSUSE Leap 16 can use FDE (LUKS + a unique btrfs partition) combined with TPM auto-unlocking (a simple box to tick during installation) with only a separate EFI partition. Why did Fedora do things differently?
I’m not an expert but I’m curious to learn the reasons.
But figured that can’t be where things are stored (FreeBSD only has 1.3MB usage there ). But if /boot can be on the same drive as /, I wonder why separate partitions is default?
I guess it gets into subvolumes and separate /home stuff as a benefit, but the simple traditional large / avoids tiny /boot issues
(open)SUSE has a patch for GRUB to put grubenv into the Btrfs bootloader pad. There, GRUB is allowed to write to it. This patch is submitted upstream so Fedora should see it once the patch is merged, and Fedora rebases GRUB.
For encryption, the (open)SUSE installer has the ability to configure GRUB for encryption support, Fedora’s installer doesn’t.