why it used to be Grub?
when we can’t activate the snapshot to boot from it, that meant the boot partition is damaged. That used to be the grub.
to check it, updated all but grub - it worked.
Another time upgraded all, except of grub and linux-zen. It workrd. Then updated grub - it failed to boot, showed “out of memory” during loading initramfs. So, my guess, the new version of Grub isn’t compatible with new Nvidia drivers…
Grub has nothing to do with any drivers.
I suspect your hardware may be incompatible with the latest grub as well. You’re not the first to have trouble with it.
yes, indeed, grub has a lot to do with GPU drivers, once we talk about Nvidia as sometimes it needs to set a “GRUB_GFXMODE=****” resolution manually in grub /etc/default/grub . But in this our case the issue was about Grub damaging boot partition, so we could not recover with a snapshot either.
if Grub has issues with some hardware, that is an issue of Grub developers as they did wrong things, because previous version of Grub works great.
just as an information, what had been done. There were three runs made
The initial ISO is - garuda-mokka-linux-zen-250308.iso
a) upgrade with exclusion of “linux-zen, linux-zen-headers, and all nvidia updates”, but we upgraded grub. It worked with all Nvidia drivers exisitng in initial ISO. It worked smooth
b) then we decided to add Nvidia to “a” setup. After upgrading Nvidia, the initramfs could not load as Grub was there.
c) then we made upgrade where these weer excluded: grub, grub-btrfs. All worked
That meant - the issue is in grub as it damages the boot partition, so nothing can boot after that…
I don’t think anything is being damaged. How are you coming to that conclusion?
after upgrade it said - initramfs is out of memory, can’t load it.
Then it was feasible to boot with a snapshot . When set that snapshot as bootable one, then reboot - it said a message that can’t find any initramfs either - out of memory.
What does that mean? The issue is with boot sector, isn’t it?
the sjopshot was - before that update. So, all is there. New version of Grub makes something wrong
Can we see a screenshot/picture of this message?
don’t have them now and not going to damage the setup again, as don’t want to spend another hour to setup the system without grub update again.
look, the sequence
note - we use Nvidia proprietary drivers as need them. The open source drivers do work bad for our needs.
we upgraded all, including Grub, but excluded: linux-zen, linux-zen-headers and all nvidia updates.
updated grub worked smooth with initial kernal, initial Nvidia drivers etc
then we upgraded Nvidia - it stopped to boot by having “initramfs out of memory” with kernal panic after that
this means that we have an issue in between “new grub” and “new Nvidia drivers”.
New Nvidia drivers and new Linux-zen kernel do work normal with initial grub version.
not saying grub has issue with Nvidia drivers, but saying - grub can’t handle that kernel with Nvidia updated drivers (dkms) and something is wrong with it.
But the issue isn’t ending there. Snapshot must boot smooth, only if boot partition is normal. When we chose a snapshot and it could not boot either - this means that boot partition is damaged…
just found the photo in the internet. All was like this, including boot with a recovered snapshot
tried the install with Nvidia open source drivers. The CPU had 100% load by wayland, it was impossible to move a mouse. It took around 15-20 minutes to open a terminal and type “htop” to see what is happening. Wayland was loading all cores of CPU to 95-97%
This can mean that things had been loaded wrong.
But truth is there as we can’t boot with USB by using latest ISO of next distros: Shani, MX Linux, Manjaro, Garuda. Just tried that only.
Cache OS latest ISO boots well, but went into the same issue after installation. It was a check to see - is the issue with Garuda or Global. The issu is Global and it is in between “new Grub” that can’t handle things well
Because, when we made next
updated all, but grub, grub-btrfs - it worked smooth
when we updated grub, we saw a message after update - when each package usually leaves some messages for admins etc. It was this message -
/usr/bin/grub-probe: warning: uknown device type nvme1n1
after reboot, we got the issue - out of memory
I don’t think those exist?
Not relevant to this case.
p[pretty much relevant, as system went broke after grub update. And during update grub leaves that message.
Nvidia doesn’t make open source GPU drivers like AMD does, but Linux community has open source drivers for Nvidia - that is what we are talking about
you couldn’t be unaware of this
and Nvidia GPU open source drivers are bad, very bad. We are not gaming, just for daily use.
When 240Hz monitor runs only at 60Hz with open source drivers - that isn’t good. If you are lucky, sometimes it can run at 120Hz. With proprietary drivers it runs at 240Hz wit all of the features activated (dolby vision etc), as used to be.
Not to take over this tread but I seem to be having the same problem After the update
See enclosed screenshot
No snapshots work they all error the same way and different kernels do the same.
Things I’ve tried reinstalling grub, checked bios checking uuids but still have the same problem.
Also when booting from live usb with the new mokka iso I get the same results, if boot dragonized iso from August it boots fine into the live environment ?
After alot of searching and trial and error i was getting, it would seem that the great Ms of this world decided that i needed my TPM turned on in my bios with there updates.
@uube01 please check if TPM is turned off in your bios
offtopic
What the hell is M$ doing in my BIOS? Unbelievable.
Further text censored.
@Apocalypticus so it solved it for you too 
There was nothing to solve here. No Windows on any computer here 
TPM is off, always was
just checked another two laptops - the same hardware. On that two the TMP was enabled.
Have two working laptops and third is for browsing internet etc, checking updates etc. On that testing one and one of working laptops the TPM was enabled. But how? That feature always was disabled. After disabling TPM on that two laptops, all is working smooth. However, the new ISOs of Garuda are not booting well though…
so I actually had a kernel panic last week too. But I found that having a second kernel to switch to helped in terms of finally getting my system up to date without the chaos of what you’re having to deal with now. Make sure you have another kernel like LTS or any one of the others because that will help in a situation like that
