You are not logged in.
...
What *exactly* did you do?
In terminal I opened as a root
/etc/kernel/cmdline
root=/dev/mapper/disk rootfstype=btrfs rootflags=subvol=/@ rw quiet splash modprobe.blacklist=pcspkr zswap.enabled=0
and changed it to
/etc/kernel/cmdline
root=/dev/mapper/disk rootfstype=btrfs rootflags=subvol=/@ rw noresume modprobe.blacklist=pcspkr zswap.enabled=0
and then rebooted.
Last edited by marol (2024-03-19 22:02:42)
Offline
Offline
I'm about to give up.
I tried here everything including the information in those two links about mkinitcpio and UKIs, but problem remains. I edited /proc/cmdline according to the link, given by Head_On_a_Stick. It didn't work - arch boots with quiet splash like I didn't remove it, but I did! Now I'm thinking to edit it aagain this way:
root=/dev/mapper/disk rootfstype=btrfs rootflags=subvol=/@ rw resume=/dev/mapper/swap modprobe.blacklist=pcspkr zswap.enabled=0
But I'm not sure that it will work.
Offline
Punching random stuff into /etc/kernel/cmdline is pointless unless you manage to actually apply the changes and
I edited /proc/cmdline
is not a thing.
Did you forget to mount the ESP before you ran mkinitcpio?
lsblk -f
Offline
Punching random stuff into /etc/kernel/cmdline is pointless unless you manage to actually apply the changes and
I edited /proc/cmdline
is not a thing.
Did you forget to mount the ESP before you ran mkinitcpio?lsblk -f
No, I didn't. You can see my first post - there are fstab and lsblk -f codes.
As for mounting partitions, I recalled - according to script it goes this way:
# mount -o ssd,noatime,compress=zstd:1,space_cache=v2,autodefrag,subvol=@ /dev/mapper/cryptroot /mnt
# mount -o ssd,noatime,compress=zstd:1,space_cache=v2,autodefrag,subvol=@home,nodev /dev/mapper/cryptroot /mnt/home
# mount -o ssd,noatime,compress=zstd:1,space_cache=v2,autodefrag,subvol=@snapshots,nodev,nosuid,noexec /dev/mapper/cryptroot /mnt/.snapshots
# mount -o ssd,noatime,compress=zstd:1,space_cache=v2,autodefrag,subvol=@var_log,nodev,nosuid,noexec /dev/mapper/cryptroot /mnt/var/log
# mount -o ssd,noatime,compress=zstd:1,space_cache=v2,autodefrag,subvol=@pacman_pkgs,nodev,nosuid,noexec /dev/mapper/cryptroot /mnt/var/cache/pacman/pkg
# mount /dev/sdX1 /mnt/efi
# swapon /dev/sdX3
I did
...
# swapon /dev/sdX3
# mount /dev/sdX1 /mnt/efi
I'm sorry, but does it matter?
Offline
# mount /dev/sdX1 /mnt/efi
If this is literal, it's most certainly wrong. But according to your OP you've an /efi on nvme0n1p1
Please post the output of "mkinitcpio -v -p linux" and afterwards
ls -lR /efi
And, just for clarification, the output of "cat /proc/cmdline" does not change, regardless of what you do to /etc/kernel/cmdline (and then run mkinitcipio and reboot)?
Offline
# mount /dev/sdX1 /mnt/efi
If this is literal, it's most certainly wrong. But according to your OP you've an /efi on nvme0n1p1
I'm sorry, my bad. You are right. When I was installing archlinux, I did
# mount /dev/nvme0n1p1 /mnt/efi
Please post the output of "mkinitcpio -v -p linux"
I tried to
mkinitcpio -v -p linux
but got this
curl -F 'file=@linux.txt' https://0x0.st
curl: (26) Failed to open/read local data from file/application
and afterwards
ls -lR /efi
ls -lR /efi
/efi:
total 8
drwxr-xr-x 3 root root 4096 Mar 16 23:41 EFI
drwxr-xr-x 2 root root 4096 Mar 21 09:03 loader
/efi/EFI:
total 4
drwxr-xr-x 2 root root 4096 Mar 17 09:32 Linux
/efi/EFI/Linux:
total 201580
-rwxr-xr-x 1 root root 73938040 Mar 17 09:31 arch-linux.efi
-rwxr-xr-x 1 root root 132476536 Mar 17 09:32 arch-linux-fallback.efi
/efi/loader:
total 4
-rwxr-xr-x 1 root root 32 Mar 21 09:03 random-seed
And, just for clarification, the output of "cat /proc/cmdline" does not change, regardless of what you do to /etc/kernel/cmdline (and then run mkinitcipio and reboot)?
No, it does not. I saw that "cat /proc/cmdline" outputs the same line as it was at the beginning of installation.
Offline
mkinitcpio tries to post to 0x0.st ???
type mkinitcpio
You haven't updated the UKI since sunday morning, though.
Offline
Signing EFI binaries...
Generating EFI bundles....
✓ Signed /efi/EFI/Linux/arch-linux-fallback.efi
File has already been signed /efi/EFI/Linux/arch-linux.efi
==> Post processing done
$ curl -F 'file=@mkinitcpio.txt' https://0x0.st
curl: (26) Failed to open/read local data from file/application
Offline
That's not the output of "type mkinitcpio", you probably have some alias or run this in script or whatever that tries to upload mkinitcpio.txt but that file doesn't exist.
Offline
I don't know what happened, but now arch booted without 2 minutes delay. And without "quiet splash".
Offline
Neither do we, but from your last - unexplained - mkinitcpio post in #34 I'd check
ls -lR /efi
cat /proc/cmdline
the UKI has probably been built and udpated…
Last edited by seth (2024-03-22 07:26:37)
Offline
Neither do we, but from your last - unexplained - mkinitcpio post in #34 I'd check
ls -lR /efi cat /proc/cmdline
the UKI has probably been built and udpated…
/efi:
total 8
drwxr-xr-x 3 root root 4096 Mar 16 23:41 EFI
drwxr-xr-x 2 root root 4096 Mar 22 08:49 loader
/efi/EFI:
total 4
drwxr-xr-x 2 root root 4096 Mar 22 00:37 Linux
/efi/EFI/Linux:
total 201580
-rwxr-xr-x 1 root root 73938040 Mar 22 00:37 arch-linux.efi
-rwxr-xr-x 1 root root 132476536 Mar 22 00:37 arch-linux-fallback.efi
/efi/loader:
total 4
-rwxr-xr-x 1 root root 32 Mar 22 08:49 random-seed
cat /proc/cmdline
root=/dev/mapper/disk rootfstype=btrfs rootflags=subvol=/@ rw noresume modprobe.blacklist=pcspkr zswap.enabled=0
Last edited by marol (2024-03-22 08:30:38)
Offline
"Yes" - the problematic hibernation entry is probably still there (have a look) and what prevents it from doing anything is the "noresume" parameter.
In future, please post verbatim outputs and not random stuff you cherry-picked out of the terminal, eg. we still don't know where that 0x0.st upload approach came/comes from.
Offline
"Yes" - the problematic hibernation entry is probably still there (have a look) and what prevents it from doing anything is the "noresume" parameter.
In future, please post verbatim outputs and not random stuff you cherry-picked out of the terminal, eg. we still don't know where that 0x0.st upload approach came/comes from.
This time I did
journalctl > logs.txt
curl -F 'file=@logs.txt' https://0x0.st
Later I will checkout that output.
Offline
So you now uploaded your entire journal to 0x0.st, you don't have to check that output, it's your journal.
And it does in no way answer any of the questions above…
Offline
Thank you, seth. This is sad to me. Does it mean that "TIMEOUT" may happen at any time of booting the system?
Offline
You can always run into timeouts, but as long as the noresume parameter is there, this specific one won't hit.
If you don't know why you're recent mkinitcpio actually worked and the previous ones didn't, you'll have to start answering questions - nobody here can possibly tell you what was going on on your system.
Offline
Well, the only thing is rolling around in my mind... to reisntall system and to follow the order of partitions mounting (as I did on my laptop, where was archlnix, luks+btrfs+secure Boot). I mean
# mount -o ssd,noatime,compress=zstd:1,space_cache=v2,autodefrag,subvol=@ /dev/mapper/cryptroot /mnt
# mount -o ssd,noatime,compress=zstd:1,space_cache=v2,autodefrag,subvol=@home,nodev /dev/mapper/cryptroot /mnt/home
# mount -o ssd,noatime,compress=zstd:1,space_cache=v2,autodefrag,subvol=@snapshots,nodev,nosuid,noexec /dev/mapper/cryptroot /mnt/.snapshots
# mount -o ssd,noatime,compress=zstd:1,space_cache=v2,autodefrag,subvol=@var_log,nodev,nosuid,noexec /dev/mapper/cryptroot /mnt/var/log
# mount /dev/nmve0n1p1 /mnt/efi
# swapon /dev/nmve0n1p2
but not
# swapon /dev/nmve0n1p2
# mount /dev/nmve0n1p1 /mnt/efi
as I did in this case. Or it doesn't matter?
Offline
Wtf are you talking about??
https://deepl.com ?
Re-installing the system won't do much about the present efi variable - unless you nuke the ESP.
But you don't have to do any of this, you need tofigure how to reliably build the UKI (you now managed it *once* but apparenly not before - for still completely undisclosed reasons) to apply kernel parameters - in general and so you can control the resume behavior (noresume for now is fine and an actual "resume=…" if you ever want to use S4.
Offline
Thank you for your advice I will work with it.
Offline