Greetings,
I am using RAID to encrypt my internal system drive. The drive is fully
allocated to OpenBSD, and I followed the procedures per the FAQ. I use RAID
encryption to also encrypt several external disks. All works as documented
except for what is described below.
I am perfectly willing to send more detailed information, and realize this may
require more interaction. However before doing so I would like to eliminate
user error before I start to send out more sensitive information.
Is this panic expected behavior since I attempted to delete a system volume
currently in use? I ask this because I would expect a message for example that
the device is busy; similar behavior to umount command.
Please advise. All the work I do is on a stable OpenBSD releases.
Sincerely,
John Molloy
----------------- limited sendbug with -PD options --------------------
Synopsis: bioctl -d on encrypted system device causes kernel panic and boot
corruption
Category: kernel
Environment:
System : OpenBSD 6.0
Details : OpenBSD 6.0-stable (GENERIC.MP) #0: Sun Nov 6 09:05:53 EST
2016
[email protected]:/usr/obj/sys/GENERIC.MP
Architecture: OpenBSD.amd64
Machine : amd64
Description:
bioctl -d sd<x> where x is the encrypted system device.
Command bioctl -c C -l /dev/<duid>.a -k /dev/<duid>.d softraid0.
The encrypted disk device is mounted and all works. I am using
several encrypted disks with bioctl using this mechanism.
As I was testing additional functions, I accidentally performed
bioctl -d sd<x> where sd<x> is my encrypted system volume.
This resulted in a kernel panic and the boot sector was corrupted.
How-To-Repeat:
This behavior is repeatable and consistent.
bioctl -d <encrypted system volume>
Fix:
Booted from a external disk. repaired the encrypted system volume and
manually performed installboot
