You are not logged in.
I'm trying to setup a zfs root on a luks-encrypted partition using systemd-boot (née gummiboot) as my boot-loader. (I'm changing this over from a ext4 root system which was also using a luks-encrypted root partition.)
I have the following settings for /etc/fstab, for my relevant boot entry (/boot/loader/entries/), and zfs:
/etc/fstab:
/dev/zvol/tank/swap none swap sw 0 0arch.conf:
title Arch Linux mainline kernel
linux /vmlinuz-linux
initrd /intel-ucode.img
initrd /initramfs-linux.img
options cryptdevice=UUID=<UUID>:cryptroot zfs=tank/ROOT/arch rw acpi_osi="!Windows 2012" zfs.zfs_arc_max=6442450944(where <UUID> is a real UUID)
I have set the zfs cachefile with:
zfs set cachefile=/etc/zfs/zpool.cache tankand tank/ROOT/arch has the mountpoint set as / (i.e. zfs set mountpoint=/ tank/ROOT/arch )
and I set the bootfs as:
zpool set bootfs=tank/ROOT/arch tankBut when I boot, it asks me to enter the password for the cryptroot volume, which I do. It accepts the password, but then locks up after displaying:
:: running hook [zfs]
ZFS: Importing pool tank.
:: running late hook [lvm2]
:: running late hook [zfs]
no pool available to import
:: running cleanup hook [lvm2]
:: running cleanup hook [udev]
Freezing execution.What am I doing wrong?
Last edited by emacsomancer (2017-05-01 20:01:58)
Offline
Hi,
I'm beginner, but the execution order of those hooks is in the file `/etc/mkinitcpio.conf`.
They need to have the right order, you may want to check these...
Last edited by revo86er (2017-05-02 12:16:04)
Offline
Hi,
I'm beginner, but the execution order of those hooks is in the file `/etc/mkinitcpio.conf`.
They need to have the right order, you may want to check these...
You are quite right, I should have reported this as well.
I have tried a number of different combinations. The last one (which still didn't work) being:
HOOKS="base udev autodetect modconf block keyboard zfs filesystems"Offline
I've a luks-encrypted harddrive as well, with btrfs.
My Hooks line looks like this:
HOOKS="base udev resume autodetect modconf block encrypt filesystems keyboard btrfs" you may need to add `encrypt`?
Offline
I've a luks-encrypted harddrive as well, with btrfs.
My Hooks line looks like this:HOOKS="base udev resume autodetect modconf block encrypt filesystems keyboard btrfs"you may need to add `encrypt`?
Right, sorry, I actually have just gone to a straight zfs (w/o luks) setup just to try to cut down on the number of variables. I did have 'encrypt' in there at the beginning though.
Offline
It seems the system is locking up on or after line 47 of /usr/lib/initcpio/init which is copied into the initrd as init
run_hookfunctions 'run_cleanuphook' 'cleanup hook' $CLEANUPHOOKS
if [ "$(stat -c %D /)" = "$(stat -c %D /new_root)" ]; then
# Nothing got mounted on /new_root. This is the end, we don't know what to do anymore
# We fall back into a shell, but the shell has now PID 1
# This way, manual recovery is still possible.
err "Failed to mount the real root device."
echo "Bailing out, you are on your own. Good luck."
echo
launch_interactive_shell --exec
elif [ ! -x "/new_root${init}" ]; then
# Successfully mounted /new_root, but ${init} is missing
# The same logic as above applies
err "Root device mounted successfully, but ${init} does not exist."
echo "Bailing out, you are on your own. Good luck."
echo
launch_interactive_shell --exec
fi
if [ "${break}" = "postmount" ]; then
echo ":: Post-mount break requested, type 'exit' to resume operation"
launch_interactive_shell
fi
# this should always be the last thing we do before the switch_root.
rdlogger_stop
exec env -i \
"TERM=$TERM" \
/usr/bin/switch_root /new_root $init "$@"
# vim: set ft=sh ts=4 sw=4 et:Perhaps add some more `echo`s to determine which line exactily it freezes on?
Edit:
Added more of the code as it could progress further without producing output.
Last edited by loqs (2017-05-03 18:14:15)
Offline