#Boot failed, stuck in emergency mode with root user locked [SOLVED]
134 messages · Page 1 of 1 (latest)
I suspect that reinstalling the kernel and regenerating the initramfs would sort this out
You'll probably need your iso usb
What
@jade scarab
What's wrong
wrong ping :v)
do you have a live usb
[Reply to:](#1397627514771472425 message) installed kernel and regenerated the initramfs and it didnt work
yea
alright
then do usermod -U root && passwd root
set a password, and regenerate the initramfs
If this is systemd-boot I think the root account needs activating in a config somewhere? I've not used it but I think I read that
journalctl -xb would be useful to look at, as it mentions
funny enough really nothing stands out in journalctl
oh wait
i should prob fsck nvme0n1p1
still same outcome after regerating initramfs and changing password of root
May use kernel reinstall, as mentioned.
Yeah, I'm curious how you're regenerating the initramfs
mkinitcpio -P
Most complete way is to pacman -S <your kernel>
That should do the trick. Is your ESP mounted?
yeah
Well fiddlesticks
Does systemd-boot have a way of reinstalling it or regenerating its config?
unsure, i think pacstrap -K from when you install arch from the boot iso installs it but aside from that hmm
no
im wrong
wait 💀
im all confused here
I can only go what I know of grub so please take it all with a grain of salt
I can only figure that bootloaders would work similarly
But for grub you run grub-install against the ESP to install grub there, and then grub-mkconfig to generate the grub config it will refer to to print all the boot entries when it starts
reinstalled kernel through pacman -S linux same issue though
I'm suggesting doing whatever the systemd-boot equivalent is, to make sure everything lines up properly
Assuming it has an equivalent.
systemd-boot is tied to systemd, it comes with as a package
it makes it more complicated
thats why i like using really anything that systemd supports because in theory it'd have no issue connecting the dots from systemd->systemd-boot but something's gotta be wrong
ext4
this is my fstab file and my loader entry arch.conf
We've not really seen an error yet, besides that fsck warning
hmm looking closer, this does look quite error-y
looked it up and ppl said they could run with it on, it's just bc the nvme firmware is "bad" like the cheap kinda bad
Actually, looking it up it is looking less error-y and more just a warning
i dont get it
modules literally load
These logs look like they're from the iso
yup, chrooted into nvme0n1p2
exactly what muesli said
the error is that the initramfs fails to mount the rootfs and switch to the root
if im chrooted into /mnt does journalctl read from iso??
excuse my mistake, those are from the iso
nvidia driver i saw
and is dumping core
that is not a problem
no it's not ugh
excuse that comment aswell im chrooted
did you regenerate initramfs
yeah
something is causing it to fail to mount the real root
when chrooting?
what kernel?
here is the error
failed to mount sysroot
this is the error that causes your system to not boot
what did fsck result in
this, and a dirty bit (which i unset)
oh oh
Should the init not be fscking automagically? 
it only does it periodically
[Reply to:](#1397627514771472425 message) Should the init not be fscking automagically? 
unless you create /.autorelabel which causes it to fsck on every boot
it fsck's only when the time has arrived iirc
looking at this entry, should it be root=UUID=blahblah?
no, the fstab looks fine
Not in the fstab, in the boot entry
root=UUID=ksdjfksdj applies to the kernel boot parameters
fixed arch.conf and rebooting
where?
In the 3rd pic of the three
If it is that I'm going to be upset on your behalf @jade scarab
still no, even after the fix sysroot is failing to mount the drive
i think those are the boot parameters from the arch iso
I think you need to regen initramfs again if you didn't
damn
still, same outcome
it's gotta be efibootmgr
i def did something wrong
but like ?? it's exactly the same UUID
including the UUID=?
yeah
can you show a pic of the efibootmgr output?
no args should show what's in your nvram
I think so
🎉
nice
im writing this down in my "when my power goes out what to fix for arch breaking efi partitions" checklist
Boot failed, stuck in emergency mode with root user locked [SOLVED]
ugh
my bad guys LMAO
i usually get the efibootmgr right
If this is a single OS machine, may I suggest a unified kernel image? :D Dodge all this EFI/bootloader nonsense
"Hey compooter, run this"
unified kernel image sounds like something i need, because windows is the only boot manager that isn't fragile as my linux boot stub