nixos-23.11 aarch64 kexec kernel panics on switch_root

[XA21X]

This is on an ARM instance on Oracle Cloud. The 23.05 aarch64 kexec works but 23.11 kernel panics. I obtained this output from the serial console after changing the kernel command line to include loglevel=7 debug, but don't see a useful error message:

[    8.810943] stage-1-init: [Sun Jan 21 04:46:19 UTC 2024] starting device mapper and LVM...
[    8.883307] stage-1-init: [Sun Jan 21 04:46:19 UTC 2024] mounting tmpfs on /...
[    8.892663] stage-1-init: [Sun Jan 21 04:46:19 UTC 2024] mounting ../nix-store.squashfs on /nix/.ro-store...
[    8.893960] loop0: detected capacity change from 0 to 234904
[    8.922896] stage-1-init: [Sun Jan 21 04:46:19 UTC 2024] mounting tmpfs on /nix/.rw-store...
[    8.954467] stage-1-init: [Sun Jan 21 04:46:19 UTC 2024] mounting overlay on /nix/store...
[    9.092900] Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000100
[    9.096122] CPU: 2 PID: 1 Comm: switch_root Not tainted 6.6.12 #1-NixOS
[    9.097322] Hardware name: QEMU KVM Virtual Machine, BIOS 1.5.1 06/16/2021
[    9.098486] Call trace:
[    9.098778]  dump_backtrace+0xa0/0x128
[    9.099105]  show_stack+0x20/0x38
[    9.099356]  dump_stack_lvl+0x48/0x60
[    9.099673]  dump_stack+0x18/0x28
[    9.099981]  panic+0x318/0x390
[    9.100299]  do_exit+0x8c4/0x9f0
[    9.100603]  do_group_exit+0x3c/0xa0
[    9.101084]  __arm64_sys_exit_group+0x20/0x28
[    9.101438]  invoke_syscall+0x50/0x128
[    9.101734]  el0_svc_common.constprop.0+0x48/0xf0
[    9.102215]  do_el0_svc+0x24/0x38
[    9.102621]  el0_svc+0x48/0x138
[    9.102914]  el0t_64_sync_handler+0x100/0x130
[    9.103352]  el0t_64_sync+0x190/0x198
[    9.103631] SMP: stopping secondary CPUs
[    9.104402] Kernel Offset: 0x5ef40cc00000 from 0xffff800080000000
[    9.105550] PHYS_OFFSET: 0x40000000
[    9.105970] CPU features: 0x0,00000021,70020143,1000720b
[    9.106403] Memory Limit: none
[   15.439267] ---[ end Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000100 ]---

[Cu3PO42]

I believe I am experiencing the same issue with an aarch64 server hosted on Hetzner.

[Mic92]

Does downgrading the linux kernel help?