Dear Maintainer,

The system experienced severe NVMe I/O errors while running normally. Multiple
EXT4 partitions on the same NVMe device (nvme0n1p6 and nvme0n1p8) started
reporting write and buffer I/O errors.

The EXT4 journals subsequently aborted and both filesystems were remounted
read-only. systemd-journald was unable to write its journal, followed by a
systemd-journal segmentation fault.

The system then experienced a kernel panic:

Kernel panic - not syncing: Attempted to kill init! exitcode=0x0000000b

The affected kernel was 7.1.12+deb14-amd64.

The system was being used normally when the failure occurred. I did not perform
any unusual disk operations immediately before the failure.

The outcome was that the filesystem became read-only and the system eventually
crashed with a kernel panic, requiring a reboot. The logs also reported
potential data loss.

I expected the system to handle the NVMe I/O failure without causing PID 1 to
terminate and triggering a kernel panic.

Relevant kernel log excerpts are included below.