Hello, Another data point for this bug, in case it helps.
Two Debian 12 KVM guests (virtio-scsi, ext4 root, Docker with overlay2), running on different physical hosts, froze after upgrading to linux-image-6.1.0-55-amd64 (6.1.192-1). There were four freezes in total. Every freeze started right after an hourly cron job "sync; echo 3 > /proc/sys/vm/drop_caches": the last journal/syslog entry each time was "drop_caches: 3", at the first run after each boot (roughly 10 to 55 minutes after boot). After that, disk I/O blocked completely: journald stopped writing, processes were stuck in D state and systemd watchdog timeouts followed. The kernel itself kept running (netfilter messages on the console, ICMP replies). No oops reached the journal. On 6.1.0-54 (6.1.190-1) the same job ran hourly for weeks without problems, and it still completes normally today. Both machines are in production, so I am not able to reproduce the freeze on 6.1.192-1 to capture the oops. I am running 6.1.0-54 for now, will test 6.1.192-2 once it is available, and will report back. Thank you for the quick work on this.

