Try the latest live iso and boot it legacy (bios) mode. No reason to poke the installed system if you can’t even use the live iso.
1 Like
You can get the live iso journal too, probably has the same problem as your installed kernel.
Yes, I agree, it is a good idea to check if the live ISO runs.
This probably means, that the kernel module vfat is not loaded.
And that’s the reason why /efi cannot be mounted I guess.
Can you also give the output of lsblk -f to see which directory is your esp?
So, for now it looks to me that something went wrong with your kernel installation…
Are you using dracut or mkinitcpio?
Systemd-boot probably can’t find your new kernel because it isn’t properly installed. That would fit also the missing module.
Just out of curiosity have you tried Cashy OS? Try looking at the kernel options. Since its also an arch based system it may could help with your issue
Sorry for dropping off the radar everyone, had midterms to deal with.
I have just successfully booted the latest live disk and confirmed it was running linux 6.1.9. I can reinstall, but I’d really like to figure out what’s wrong with systemd and see if the installed kernel boots with the expanded swap file - I’ve never yet run a system on so little swap before so I’m curious about that. FYI, I have already reinstalled the linux package and that didn’t fix it, first thing I did when I noticed the problem.
lsblk -f output:
NAME FSTYPE FSVER LABEL UUID FSAVAIL FSUSE% MOUNTPOINTS
sda
├─sda1 vfat FAT32 DAFD-E999 736.8M 26% /efi
└─sda2 ext4 1.0 endeavouros 906c4ac1-6333-4bd6-8d50-61b04bb71ded 81.5G 24% /
I am currently using mkinitcpio, I tried to install the gallium-os kernel to fix this and it changed it. ran into some error installing the kernel, but it seemed to install that just fine (related to systemd problem?)
I have not tried Cashy OS or even heard of it, what’s that like?
Okay. Can you try and reinstall sudo pacman -S dkms linux linux-headers. In order to get the packages downloaded freshly you would have to remove the latest versions from the pacman cache.
If you are using the zen-kernel replace the linux and linux-headers packages in this line with the zen equivalent.
Please give us then the output of the installation process and check if the mkinitcpio-hook is run without errors.
How did you install it? Is that available in AUR?
I guess you refer to systemd-boot. I assume that there is nothing wrong with it. The problem appears to be the kernel.
Can you give the output of lsmod from the emergency mode of the 6.2-kernel?
I assume you installed the 6.2-kernel by just using sudo pacman -S linux linux-headers or by a regular system update, right?
Output of install command:
(3/3) checking package integrity [------------------------------------] 100%
(3/3) loading package files [------------------------------------] 100%
(3/3) checking for file conflicts [------------------------------------] 100%
:: Processing package changes...
(1/3) installing dkms [------------------------------------] 100%
Optional dependencies for dkms
linux-headers: build modules against the Arch kernel [installed]
linux-lts-headers: build modules against the LTS kernel [installed]
linux-zen-headers: build modules against the ZEN kernel
linux-hardened-headers: build modules against the HARDENED kernel
(2/3) installing linux [------------------------------------] 100%
Optional dependencies for linux
wireless-regdb: to set the correct wireless channels of your country
linux-firmware: firmware images needed for some devices [installed]
(3/3) reinstalling linux-headers [------------------------------------] 100%
:: Running post-transaction hooks...
(1/5) Arming ConditionNeedsUpdate...
(2/5) Updating module dependencies...
(3/5) Install DKMS modules
(4/5) Updating linux initcpios...
==> Building image from preset: /etc/mkinitcpio.d/linux.preset: 'default'
-> -k /boot/vmlinuz-linux -c /etc/mkinitcpio.conf -g /boot/initramfs-linux.img
==> Starting build: 6.2.2-arch1-1
-> Running build hook: [base]
-> Running build hook: [udev]
-> Running build hook: [autodetect]
-> Running build hook: [modconf]
-> Running build hook: [kms]
-> Running build hook: [keyboard]
==> WARNING: Possibly missing firmware for module: xhci_pci
-> Running build hook: [keymap]
-> Running build hook: [consolefont]
==> WARNING: consolefont: no font found in configuration
-> Running build hook: [block]
-> Running build hook: [filesystems]
-> Running build hook: [fsck]
==> Generating module dependencies
==> Creating zstd-compressed initcpio image: /boot/initramfs-linux.img
==> Image generation successful
==> Building image from preset: /etc/mkinitcpio.d/linux.preset: 'fallback'
-> -k /boot/vmlinuz-linux -c /etc/mkinitcpio.conf -g /boot/initramfs-linux-fallback.img -S autodetect
==> Starting build: 6.2.2-arch1-1
-> Running build hook: [base]
-> Running build hook: [udev]
-> Running build hook: [modconf]
-> Running build hook: [kms]
==> WARNING: Possibly missing firmware for module: ast
-> Running build hook: [keyboard]
==> WARNING: Possibly missing firmware for module: xhci_pci
-> Running build hook: [keymap]
-> Running build hook: [consolefont]
==> WARNING: consolefont: no font found in configuration
-> Running build hook: [block]
==> WARNING: Possibly missing firmware for module: qed
==> WARNING: Possibly missing firmware for module: qla1280
==> WARNING: Possibly missing firmware for module: wd719x
==> WARNING: Possibly missing firmware for module: bfa
==> WARNING: Possibly missing firmware for module: qla2xxx
==> WARNING: Possibly missing firmware for module: aic94xx
-> Running build hook: [filesystems]
-> Running build hook: [fsck]
==> Generating module dependencies
==> Creating zstd-compressed initcpio image: /boot/initramfs-linux-fallback.img
==> Image generation successful
(5/5) Checking which packages need to be rebuilt
GalliumOS kernel; I just ran yay -S galliumOS. I believe I may have copied the package name from the arch wiki. I don’t remember exactly what happened but the install of the kernel itself failed, I never actually had that kernel on the system. But it changed dracut to mkinitcpio before it tried to install and nothing broke so I just left it.
6.2 did install from a regular system update, yes.
I’ll post the lsmod momentarily, but I want to repeat something first because it seems to have fallen by the wayside. I do not have linux-zen installed. I had installed it to see if it could boot, but I uninstalled it nearly a week and a half ago now. my systemd boot screen continues to show it. I’ve verified that there are no linux-zen files in my boot folder. I don’t know why it still shows up in the boot menu but I can find no evidence to indicate this is a problem with what’s installed or not installed on the system. (for the record, despite this install command I just ran, it also still says linux is 6.2.1, not 6.2.2.)
Now, in my past experience with grub (this is my first system running systemd), the bootloader has a config file that needs to be regenerated after installing kernels. I’ve seen no such thing in the systemd documentation, but is there nothing else that might have not gotten refreshed in some way?
Lsmod output from the emergency boot:
Module Size Used by
ext4 1163264 1
crc32c_generic 16384 0
crc16 16384 1 ext4
mbcache 16384 1 ext4
jbd2 217088 1 ext4
crct10dif_pclmul 16384 1
crc32_pclmul 16384 0
crc32c_intel 24576 2
polyval_clmulni 16384 0
polyval_generic 16384 1 polyval_clmulni
gf128mul 16384 1 polyval_generic
serio_raw 20480 0
ghash_clmulni_intel 16384 0
atkbd 40960 0
sha512_ssse3 57344 0
libps2 20480 1 atkbd
aesni_intel 401408 0
vivaldi_fmap 16384 1 atkbd
crypto_simd 16384 1 aesni_intel
cryptd 32768 2 crypto_simd,ghash_clmulni_intel
i2c_hid_acpi 16384 0
xhci_pci 24576 0
i2c_hid 40960 1 i2c_hid_acpi
xhci_pci_renesas 24576 1 xhci_pci
i8042 53248 0
video 73728 0
serio 28672 4 serio_raw,atkbd,i8042
wmi 45056 1 video
pkcs8_key_parser 16384 0
crypto_user 24576 0
Was that before your problems started, so before updating to the new kernel?
The installation itself looks fine.
For manually updating systemd-boot run sudo bootctl update.
Please also check the auto-update service for systemd-boot with systemctl status systemd-boot-update.service
(Side note: I use grub, therefore I don’t have experience with systemd-boot, but as from your installation output I don’t see any pacman-hook updating the bootloader I guess this service is running.)
That means that you installed the package group galliumos from the aur. Judging by the galliumos-kernel-package: https://aur.archlinux.org/packages/linux-galliumos it’s out-of-date since 2019.
This list seems rather short and very essential modules, as e.g. fat or vfat are not loaded. As suspected, I think the main problem here is that quite a few needed kernel modules are not loaded. It also seems likely that the problem was somehow caused by the switch from dracut to mkinitcpio.
Just for comparison you can boot a proper working kernel and run lsmod again. Then you will probably see a difference in the (number of) loaded kernel modules.
For comparison: In the lsmod-output you posted are 29 modules, if I run lsmod on my very basic laptop (no special hardware, just intel graphics, etc.) I get 189 loaded kernel modules.
If the problems with the failed kernel installation/boot appeared after the installation of GalliumOS, I think you might have some very out of date configurations on your machine now.
I believe my attempt to install galliumOS was prior to the 6.2.2 update, yes. That would also explain why the install failed. Do you know how to update these configurations?
sudo bootctl update says that the [bootloader].efi file is already in place. It doesn’t seem to be doing anything with regards to what kernels I have or don’t have installed.
Would reinstalling systemd possibly work? And would I likely break my system trying?
Hmm… I guess one would have to check in depth what the Gallium configuration actually did. Also you won’t find official wiki documentation on it as it’s from AUR.
I think my knowledge about mkinitcpio and kernel modules is not good enough to figure that out in a decent time.
I assume you could try to switch back to dracut and reinstall kernel and headers. But there might still be problems afterwards.
Honestly: If it would be my system, I think I would just do a data backup and a clean new EndeavourOS-install from the newest ISO.
Systemd is the init-system that is responsible for a lot of things in the system.
If you mean to reinstall systemd-boot, I think that won’t solve the kernel freeze issue. I don’t even think that systemd-boot is broken. To me it seems more connected to kernel modules/mkinitcpio
1 Like
Understood. At this point I’m honestly thinking the same, I’m just annoyed I can’t fix this. There’s a thread on the arch forums with a very similar/identical error, but the given commands are maddeningly obtuse and the documentation isn’t much help.
I can link it if anyone is curious but I’m probably just going to reinstall and see if making the swapfile larger keeps the newer kernel running.
1 Like
Feel free to share the arch forum link. Maybe it will help someone who is searching for a solution for this error.
Please let us know if it’s running after the fresh install.
Here’s the forum link… https://bbs.archlinux.org/viewtopic.php?id=152284
I’m looking in /boot (EFI partition, right?) and all the right img files are there. Meanwhile, the efibootmgr command at the bottom is poorly explained and not much better on the arch wiki.
This is now relevant because the installer freezes up after a minute or so of running. I’d assume the default size swapfile is filling up, and it’s just taking longer because the installer is running less stuff than the full install? Seems logical to me, anyway. I have no idea how to change the swapfile size on a live boot from outside… Can I chroot into a live disk from a main install?
When the thread from the arch forum has been created systemd-boot was still called gummiboot and not yet part of systemd. So, that’s probably not very suitable for your software.
But yes, if the images are in /boot, systemd-boot should find them.
Are you doing net or online install? I had the same problem with my quite weak hardware during my attempted net install and just as you said, I think I ran out of memory. What I ended up doing was that I just did an offline installation. If you want to switch DEs, that would be doable afterwards as well.
If you are more adventurous, you could also do a “blank” arch install, include the endeavouros repo and install everything manually. It is possible to do that with the guide from the arch wiki, but it takes a lot of time and work to get everything running the way you want it (that’s the reason I am using endeavour).
I doubt you can change the swap size of the live-ISO, but I wouldn’t bet on it.
I would suggest to you to install a new system, it will be done faster, but before you install, turn off secure boot, it will make your life simpler. Use the default partitioning if you don’t want learn how to create an efi partition. Also you don’t need an efi partition if you use legacy boot, it should be supported on older machines.
Maybe use grub instead of systemd-boot, it has more configuration guides for newbies.
Keep it simple.
I’m doing an install from a USB, not net. As for offline/online, I can’t even launch the calamares installer before it freezes.
I tried doing a blank arch install and found it wasn’t really worth the hassle, but maybe I’ll try again. If I get as much of a toxically elitist response on the arch forums as I did last time, it may turn me off vanilla arch entirely.
EDIT: I just realized I can probably use the old live disk (running 5.15) to install the system and then update afterward. I didn’t delete the iso when I downloaded the latest. I can’t actually do anything with this until tomorrow but I’ll try as soon as I can.
Secure boot is already off, I can’t boot live disks on this machine if it isn’t. I know how to set up my own partitions and it seems doing the default partitioning is what got me into this pickle in the first place. As for bootloaders, at this exact moment I can’t even run the installer without it freezing, so unless legacy mode also means the system is dumping less to the oddly small swapfile then I’m not quite there yet.
I would just create a new live USB from the latest ISO. Then reinstall using erase disc with swap file and select the bootloader you want. Either systemd-boot or grub. Secure boot off, CSM disabled.
Edit: The method used to create the live usb also can cause you problems.