[SOLVED] Failed to start Load Kernel Modules

n8jadams · 2022-07-09 02:54:16

Hi, I did some moving around drives (with a gparted live volume) and repartitioning/copying stuff, and finally got my EFI bootloader to show the Arch linux option.

The only problem is, I'm getting this error:

[FAILED] Failed to start Load Kernel Modules.
[FAILED] Failed to mount /boot.
[DEPEND] Dependency failed for Local File Systems.
[FAILED] Failed to mount /c.
[FAILED] Failed to mount /d.
You are in emergency mode...

My /etc/fstab looks like this

# /dev/sda3 LABEL=archlinux
UUID=924bc4b4-47c6-4264-b17b-a9d854b819fc	/         	ext4      	rw,relatime	0 1

# /dev/sda2 LABEL=Windows10
UUID=60306AE7306AC41E	/c        	ntfs      	rw,nosuid,nodev,user_id=0,group_id=0,allow_other,blksize=4096	0 0

# /dev/sdb1 LABEL=CRUCIAL\134x202TB
UUID=24A21F40A21F1640	/d        	ntfs      	rw,nosuid,nodev,user_id=0,group_id=0,allow_other,blksize=4096	0 0

# /dev/sdb2 LABEL=EFI
UUID=67E3-17ED      	/boot     	vfat      	rw,relatime,fmask=0022,dmask=0022,codepage=437,iocharset=ascii,shortname=mixed,utf8,errors=remount-ro	0 2

# /dev/sda4
UUID=cb5c0b12-4a3d-46a2-bb16-4040d75bdeff	none      	swap      	defaults  	0 0

I can boot into my arch live usb volume just fine, and everything looks normal, the .img file is in place, the partition UUIDs are correct in my Clover configuration xml file. (This was a hackintosh machine, I nuked the hackintosh partition but kept the Clover EFI bootloader)

What else can I do to figure out what's going on?

Last edited by n8jadams (2022-07-10 15:27:47)

Trilby · 2022-07-09 03:43:41

If you repartitioned, the UUIDs chage. You say they are correct in your clover config, but that's not relevant as the kernel is booting. What is relevant is your fstab. Old UUIDs there would explain all your symptoms.

n8jadams · 2022-07-09 14:58:41

I mounted all of my drives while using the live install volume. Wouldn't the genfstab command have accounted for this?

Scimmia · 2022-07-09 15:06:55

With the 'Failed to start Load Kernel Modules' in there, this almost certainly means your bootloader is loading a different kernel than what's installed. Did you update the system with the /boot partition mounted?

seth · 2022-07-09 15:07:37

n8jadams wrote:

I mounted all of my drives while using the live install volume. Wouldn't the genfstab command have accounted for this?

Depends on when where and how you ran it.
Compare your fstab to "lsblk -f"

Last edited by seth (2022-07-09 15:08:14)

ewaller · 2022-07-09 15:11:50

Scimmia's point seems relevant.
Compare the version numbers reported by uname -a and by pacman -Q linux (assuming you are running the stock kernel)

Trilby · 2022-07-09 15:14:17

n8jadams wrote:

I mounted all of my drives while using the live install volume. Wouldn't the genfstab command have accounted for this?

Live install medium, or the gparted live medium? If you only ran genfstab with the initial install, then later used gparted to repartition and "copy stuff" as you said in your first post, then the UUIDs have changed and will need to be updated.

n8jadams · 2022-07-09 16:26:12

While booted into my arch installation live usb volume, after running arch-chroot /mnt I ran this:

$ uname -a
Linux archiso 5.12.13-arch1-2 #1 SMP PREEMPT Fri, 25 Jun 2021 22:56:51 +0000 x86_64 GNU/Linux

$ pacman -Q linux
error: package 'linux' was not found

$ pacman -Q linux-zen
linux-zen 5.18.7.zen1-1

The versions are different, but I'm not exactly sure where to go from here...

Trilby · 2022-07-09 16:28:24

The uname output there is irrelevant as that's the kernel from the live usb. Use the following:

# from within the chroot
file /boot/vmlinuz-linux*
# or outside the chroot but w/ partitions mounted under /mnt
file /mnt/boot/vmlinux-linux*

Last edited by Trilby (2022-07-09 16:28:41)

seth · 2022-07-09 16:33:08

Use the following:

That's not gonna help if the OP is booting the wrong device.
"uname -a" will also work from the rescue shell.

n8jadams · 2022-07-09 16:47:14

$ file /boot/vmlinuz-linux*
/boot/vmlinuz-linux-zen: Linux kernel x86 boot executable bzImage, version 5.18.9-zen1-1-zen (linux-zen@archlinux) #1 ZEN SMP PREEMPT_DYNAMIC Sat, 02 Jul 2022 21:03:08 +0000, RO-rootFS, swap_dev 0XA, Normal VGA

So pacman said 5.18.7.zen-1-1 and the file command says 5.18.9-zen-1-zen. What should I do now?

Last edited by n8jadams (2022-07-09 16:51:16)

seth · 2022-07-09 16:54:34

Is that /boot on sdb2 or sda3?

n8jadams · 2022-07-09 17:01:03

seth wrote:

Is that /boot on sdb2 or sda3?

sdb2, sda3 is my linux (root) partition

Last edited by n8jadams (2022-07-09 17:02:23)

seth · 2022-07-09 17:07:19

Is there a /boot/vmlinuz-linux* on sda3 as well (and is that 5.18.7.zen-1-1)?

Sidebar: how did you end up w/ a dated package but updated actual kernel? Did you at some point downgrade the kernel?

n8jadams · 2022-07-09 17:56:39

To describe what I did, from my arch linux install live volume, I ran this

$ mount /dev/sdba3 /mnt
$ mount /dev/sdb2 /mnt/boot
$ arch-chroot /mnt
# and then the rest of the commands...
$ uname -a
$ pacman -Q linux-zen
$ file /boot/vmlinuz-linux*

And that's what I'm seeing.

ewaller · 2022-07-09 19:50:41

As one would expect you are using the kernel from the install media in your chroot.

What happens when you boot your system directly? The next steps to solve this depend on the answer

This depends on the kernel that is installed in your boot volume and which kernel your bootloader picks. When that kernel, whatever it is, starts it will start to look for its kernels in your root partition. If the kernel from which you are booting is different than that from which is installed on your root volume, the kernel will not find its modules. This can happen a few ways. The usual cause is for the boot partition to have not been mounted on /boot when the kernel was installed. Another cause is that the boot loader is finding a kernel other than the one you would expect -- sometimes this happens if you have two boot partitions and are not using the one you think you are using.

n8jadams · 2022-07-09 21:37:04

I can't seem to boot my system directly, it doesn't show up in my motherboard's boot menu. It shows up in the EFI bootloader, but when I boot into it, I get the original error message. Maybe I need to just nuke this partition and reinstall arch altogether, and just copy over my data once I get that working.

seth · 2022-07-09 22:14:44

Doesn't the original error drop you into a rescue shell?

seth wrote:

Is there a /boot/vmlinuz-linux* on sda3 as well (and is that 5.18.7.zen-1-1)?

Boot the iso, just mount sda3 to /mnt and check /mnt/boot/vmlinuz*

Trilby · 2022-07-10 01:34:10

seth wrote:

Doesn't the original error drop you into a rescue shell?

I can't speak for the OP's specific situation, but if modules can't be loaded some hardware will not function, and the keyboard is a common victim.

n8jadams · 2022-07-10 03:38:28

seth wrote:

Doesn't the original error drop you into a rescue shell?
seth wrote:
Is there a /boot/vmlinuz-linux* on sda3 as well (and is that 5.18.7.zen-1-1)?
Boot the iso, just mount sda3 to /mnt and check /mnt/boot/vmlinuz*

Yes, the original error drops me into a rescue shell. I can't get the keyboard to type anything there though...

$ mount /dev/sda3 /mnt
$ file /mnt/boot/vmlinux*
zsh: no matches found: /mnt/boot/vmlinuz*

Last edited by n8jadams (2022-07-10 03:39:03)

seth · 2022-07-10 07:07:23

$ file /mnt/boot/vmlinux*
zsh: no matches found: /mnt/boot/vmlinuz*

If you made the typo in the "correct" location, there's no kernel on the root partition.
So whatever you did you get the kernel ahead of the installed package, boot the installation iso and update the kernel w/ the boot partition mounted.

$ mount /dev/sdba3 /mnt
$ mount /dev/sdb2 /mnt/boot
$ arch-chroot /mnt
$ pacman -Syu linux-zen

n8jadams · 2022-07-10 15:27:31

I had some weird stuff happen

$ mount /dev/sda3 /mnt
$ mount /dev/sdb2 /mnt/boot
$ arch-chroot /mnt
$ pacman -Syu linux-zen
...
error: failed to prepare transaction (could not satisfy dependencies)
:: installing llvm-libs (14.0.6-1) breaks dependency 'llvm-libs<14.0.0' required by reflective-rapidjson
$ pacman -Rns reflective-rapidjson
...
$ pacman -Syu linux-zen
...
$ exit
$ reboot

And now it's all working. Thank you so much!

Arch Linux

#1 2022-07-09 02:54:16

[SOLVED] Failed to start Load Kernel Modules

#2 2022-07-09 03:43:41

Re: [SOLVED] Failed to start Load Kernel Modules

#3 2022-07-09 14:58:41

Re: [SOLVED] Failed to start Load Kernel Modules

#4 2022-07-09 15:06:55

Re: [SOLVED] Failed to start Load Kernel Modules

#5 2022-07-09 15:07:37

Re: [SOLVED] Failed to start Load Kernel Modules

#6 2022-07-09 15:11:50

Re: [SOLVED] Failed to start Load Kernel Modules

#7 2022-07-09 15:14:17

Re: [SOLVED] Failed to start Load Kernel Modules

#8 2022-07-09 16:26:12

Re: [SOLVED] Failed to start Load Kernel Modules

#9 2022-07-09 16:28:24

Re: [SOLVED] Failed to start Load Kernel Modules

#10 2022-07-09 16:33:08

Re: [SOLVED] Failed to start Load Kernel Modules

#11 2022-07-09 16:47:14

Re: [SOLVED] Failed to start Load Kernel Modules

#12 2022-07-09 16:54:34

Re: [SOLVED] Failed to start Load Kernel Modules

#13 2022-07-09 17:01:03

Re: [SOLVED] Failed to start Load Kernel Modules

#14 2022-07-09 17:07:19

Re: [SOLVED] Failed to start Load Kernel Modules

#15 2022-07-09 17:56:39

Re: [SOLVED] Failed to start Load Kernel Modules

#16 2022-07-09 19:50:41

Re: [SOLVED] Failed to start Load Kernel Modules

#17 2022-07-09 21:37:04

Re: [SOLVED] Failed to start Load Kernel Modules

#18 2022-07-09 22:14:44

Re: [SOLVED] Failed to start Load Kernel Modules

#19 2022-07-10 01:34:10

Re: [SOLVED] Failed to start Load Kernel Modules

#20 2022-07-10 03:38:28

Re: [SOLVED] Failed to start Load Kernel Modules

#21 2022-07-10 07:07:23

Re: [SOLVED] Failed to start Load Kernel Modules

#22 2022-07-10 15:27:31

Re: [SOLVED] Failed to start Load Kernel Modules

Board footer