I find the thread slightly confusing, as you jump from one box which seems to be dualboot to another, and it also seems that there are different issues at play here.
Sorry, I apologize for that. I apologize in advance as this is going to be a bit of a read to try to make it clear.
1. When I opened this thread. I had the problem on one computer/laptop which was dual boot. The lead system is Garuda.
I installed Mageia 2nd and I told it to install grub to partition of install. (Later in the thread it has been mentioned that Mageia Grub will not work unless in EFI. I checked the box not to touch EFI/MBR)
2. While we were working on the primary issue, updates are resulting in a message and inability to boot kernels.
The error message is as follows:
- Code: Select all
[ 0.978687] pci 0000:00:07.0: DPC: RP PIO log size 0 is invalid
/init: line 153: /user/lib/systemd-udevd: Too many levels of symbolic links
failed to send request to update environment: No such file or directory
I check the forum if I have time at work and my work PC has Mageia on it. It has Mageia Grub as the main and is just Mageia.
I decided to go ahead and update that machine. That work machine is was running 5.10.75 kernel.
After the update. I have gotten the same message I was getting with the first machine.
3. I think the thread got a little confusing because issues were trying to be solved concerning the dual boot with Garuda and Mageia.
But the problem is, it is happening where there is just Mageia and Mageia grub now. With help in this thread I achieved chainload of Mageia Grub in the first machine from Garuda grub. I got a Sparse error:
- Code: Select all
error: sparse file not allowed.
Press any key to continue...
This Sparse error is in Mageia Grub and only in Mageia grub when chainloaded. I am unable to boot any kernel in that dual boot situation from chainload. Always getting the Sparse error.
I can still boot Mageia 5.10.62 from Garuda grub in first machine with dual boot. It was suggested that I delete some kernels and I did but nothing changed. I can only boot 5.10.62 kernel. (I updated it to 5.15.4 and deleted 5.10.75 and 5.10.78)
I think that it is thought the cause is the dual boot situation.
Since I have the problem on another machine not dual booted with no odd parameters, using Mageia Grub it seems this is a bug of some sort.
I was suggesting yesterday, Sunday that we can focus on my work machine to figure out what is going on. The exact same problem and there's no dual boot or other parameters to confuse matters. For example Garuda uses btrfs which has it's own issues in interaction with ext4 installs it seems.
4. Another comment I made Sunday referenced a 3rd & 4th machine. Well just a third. As far as the 4th is concerned I will not update it at this time. It is my primary machine. It needed a repair and I just got it back up. Updating it and winding up with this problem would be a bit of a mess for me at this time. So no updates until this is sorted.
The 3rd machine is an AMD hardware laptop. Everything else is Intel. I updated Mageia on the AMD laptop. I did net install. Kernel was 5.10.62 starting.
I updated and it was supposed to be kernel 5.15.4. After the update I can't boot 5.10.62 anymore and, it seems that 5.15.4 is not even in the grub menu?
That is a dual boot and I can still boot Debian fine and it has several kernels. Just can't boot Mageia even safe mode.
Sorry this is a long read but I hope it can clearly define what is the problem.
Every time I update Mageia currently I end up with this message duplicated across 3 different PCs and 2 with Mageia Grub in lead.
- Code: Select all
[ 0.978687] pci 0000:00:07.0: DPC: RP PIO log size 0 is invalid
/init: line 153: /user/lib/systemd-udevd: Too many levels of symbolic links
failed to send request to update environment: No such file or directory
With one unable to boot there's one we can work with which is my work machine. I had to drop back to 5.10.70 which will boot without the above message.
Garuda machine is still available for now, but I thought let's not mess with it because the btrfs issues are maybe moving off the focus of the primary issue. This too many links message and the directory not existing. Unable to boot kernels after updates recently.