Context:
I updated my system last night (EndeavourOS) and it looks like the kernal didn’t update correctly. When I restarted the system and entered my password for the encrypted drive, I get an error:
[FAILED] Failed to mount /efi
See 'systemctl status efi.mount` for details.
I can’t remember the commands I used last night but I was able to check the version of the kernel I am using currently - uname -r I believe - and what is installed. There was a difference in versions.
Trying to fix the problem:
I attempted to chroot into the system via a live USB - tutourial, arch bbs & arch wiki.
However, when trying to mount the drive (/dev/sda2) I get an error message: mount: /rescue: unknown filesystem type 'crypto_LIKS'. I tried using cryptsetup luksOpen’ and ‘udisksctl unlock -b’ but both return a similar error saying it is not an encrypted device. See fdisk -l results below:
[liveuser@eos-2024.04.20 ~]$ sudo fdisk -l
Disk /dev/sda: 238.47 GiB, 256060514304 bytes, 500118192 sectors
Disk model: TOSHIBA KSG60ZMV
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 4096 bytes
I/O size (minimum/optimal): 4096 bytes / 4096 bytes
Disklabel type: gpt
Disk identifier: FC41E181-15E3-4444-8240-E68D52AFD07E
Device Start End Sectors Size Type
/dev/sda1 4096 2052095 2048000 1000M EFI System
/dev/sda2 2052096 481648511 479596416 228.7G Linux filesystem
/dev/sda3 481648512 500103449 18454938 8.8G Linux filesystem
Disk /dev/sdb: 57.3 GiB, 61524148224 bytes, 120164352 sectors
Disk model: SanDisk 3.2Gen1
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disklabel type: dos
Disk identifier: 0x7498467c
Device Boot Start End Sectors Size Id Type
/dev/sdb1 * 64 5249887 5249824 2.5G 0 Empty
/dev/sdb2 5249888 5575519 325632 159M ef EFI (FAT-12/16/32)
Disk /dev/loop0: 2.35 GiB, 2520530944 bytes, 4922912 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Snapper Snapshots:
I recently setup snapshots with Snapper since I’m using BTRFS. From what I understand, I can just roll back my system to before the system update (it takes a snapshot before and after installing anything) but I got confused on how to do that last night - troubleshooting at 2AM with a lack of sleep will do that…
What is the best way forward? I’m happy to provide more information if it helps.
EDIT: Output of lsblk
[liveuser@eos-2024.04.20 ~]$ lsblk -f
NAME FSTYPE FSVER LABEL UUID FSAVAIL FSUSE% MOUNTPOINTS
loop0 squashfs 4.0 0 100% /run/archiso/airootfs
sda
├─sda1 vfat FAT32 0BC7-CF22
├─sda2 crypto_LUKS 2 5c6d5430-3706-48e8-bffb-f680d8c19dda
└─sda3 crypto_LUKS 2 81a912d5-fb81-40ed-a60f-0af27314b661
sdb iso9660 Joliet Extension EOS_202404 2024-04-20-15-57-10-00
├─sdb1 iso9660 Joliet Extension EOS_202404 2024-04-20-15-57-10-00 0 100% /run/archiso/bootmnt
└─sdb2 vfat FAT16 ARCHISO_EFI 7156-9697
EDIT 2:
[liveuser@eos-2024.04.20 ~]$ lsblk -a
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINTS
loop0 7:0 0 2.3G 1 loop /run/archiso/airootfs
sda 8:0 0 238.5G 0 disk
├─sda1 8:1 0 1000M 0 part
├─sda2 8:2 0 228.7G 0 part
└─sda3 8:3 0 8.8G 0 part
sdb 8:16 1 57.3G 0 disk
├─sdb1 8:17 1 2.5G 0 part /run/archiso/bootmnt
└─sdb2 8:18 1 159M 0 part
# /etc/fstab: static file system information.
#
# Use 'blkid' to print the universally unique identifier for a device; this may
# be used with UUID= as a more robust way to name devices that works even if
# disks are added and removed. See fstab(5).
#
# <file system> <mount point> <type> <options> <dump> <pass>
UUID=0BC7-CF22 /efi vfat fmask=0137,dmask=0027 0 2
/dev/mapper/luks-5c6d5430-3706-48e8-bffb-f680d8c19dda / btrfs subvol=/@,noatime,compress=zstd 0 0
/dev/mapper/luks-5c6d5430-3706-48e8-bffb-f680d8c19dda /home btrfs subvol=/@home,noatime,compress=zstd 0 0
/dev/mapper/luks-5c6d5430-3706-48e8-bffb-f680d8c19dda /var/cache btrfs subvol=/@cache,noatime,compress=zstd 0 0
/dev/mapper/luks-5c6d5430-3706-48e8-bffb-f680d8c19dda /var/log btrfs subvol=/@log,noatime,compress=zstd 0 0
/dev/mapper/luks-81a912d5-fb81-40ed-a60f-0af27314b661 swap swap defaults 0 0
tmpfs /tmp tmpfs defaults,noatime,mode=1777 0 0
EDIT 3:
I think I have fixed it. I have chrooted and am busy running sudo pacman -Syu
EDIT 4: /efi still fails to mount.
Are you able to manually mount the EFI partition while chrooted? Running an update won’t fix the problem if you don’t have the EFI partition mounted properly. To that end, are you able to manually mount the EFI partition at all? That would be a very big problem if not.
While chrooted, try the following before updating:
I’m a bit confused by the partitioning scheme, as I don’t see a /boot mount, are the files usually stored in /boot stored in /efi here? Usually the EFI partition only contains EFI binaries. Unless perhaps your /dev/sda3 is a LUKS encrypted boot partition? Sorry, I’m not very familiar with Arch, it’s strange to me to not see a /boot mountpoint in the /etc/fstab
It may be a good idea to just roll back the snapshot if none of this works, though that will only change the btrfs partition, so it won’t fix any issues inside the EFI partition if there are any issues there. Here’s an article about how to do that with btrfs.
It was causing me too much of a headache to try troubleshoot and fix that I decided to wipe the drive. I’ve got Fedora Silverblue running on the machine now. Thanks for the help!
Great choice, I’m running Kinoite (Fedora Atomic KDE) myself! The atomic nature of the distros make them less prone to breakage, and much easier to recover from (since you can just roll back to the previous branch instead of restoring a BTRFS snapshot).