[RESPONDED] Fedora 39 BTRFS IO Failure errno=-5

Hello folks,

I’m running into troubles with my AMD setup.
After a suspend period, I haven’t been able to use my laptop anymore.
During the boot phase, it just freeze after trying to start GDM.
I managed to access to journalctl and I found an error with the Filesystem which seems to be corrupted now.
Did you folks experience this too?

Many thanks

It would really help if you included hardware details, adn distro used. Based on this looks like you were using LUKS encryption without setting up your suspend to be compatible. Suspend generally does not work with LUKS out of the box. You have to configure it. My guess is you did not.

1 Like

The title says Fedora 39

Hi @Damien_Bravin ,

Is this a typical Fedora 39 installation? no custom configs during installation phase?

Fedora 39 has a bunch of spins and official releases. It helps to know exactly which one, particulalry when we move away from the default Workstation, and assuming it is the one being used often leads to troubleshooting issues down the road when they are not using what we assumed they were.

I think this is why I choose to stay away from Btrfs, and since ZFS is not in the Linux kernel tree, I am currently keeping an eye on Bcachefs.

Hey, thanks for you interest in my problem folks!
I installed the « default » fedora 39 workstation.
I ve been using fedora for 15 years. I installed btrfs on my last 4 setups. I never configured anything specific for the suspend function. I would assume that it’s taken care by the default installation process.
The issue happened I updated the os with the last kernel. I aslo noticed before having the issue that I couldn’t use the second usb c port while the first one was used, but of course, I can’t say it’s related.

Have you tried the fedora live boot yet, just to check if second usb c port comes back functioning or not?

Could be something similar to this:

I also ran into this issue and my btrfs partition was completely broken. I managed to recover from backups and some data recovery but it wasn’t fun.
Looks like kernel version 6.6 and above are triggering something weird…
(I reinstalled Ubuntu 22.04 and I’m currently sticking to oem kernel 6.5).
Just to be clear: this isn’t btrfs fault, it’s a data corruption caused by something else (AMD-Vi?)

I get your point.
Indeed it could be related. I guess we’ll need some more feedback to understand the root cause.
For now I’m stuck, I tried to repair the partition using btrfsck --init-extent-tree but now I have another issue at boot time…
I’ll try with a usb boot and see if I can get back some data. Then I’ll probably format and reinstall.
I’m in a period of my life where I needed the laptop to just work and I can’t spend hours on this :frowning: