Panic Freebsd 9.1 using "export jail_plex_geek_lan_zfs_datasets"

[Miguel Clara]

I'm trying to get a zfs dataset visible inside a jail

But when I used the flag:

export jail_plex_geek_lan_zfs_datasets="tank/Downloads"

And restart the jail the server rebooted, in the next boot I can see then
panic:

Script /usr/local/etc/rc.d/ezjail running


Fatal trap 12: page fault while in kernel mode
cpuid = 0; apic id = 00
fault virtual address   = 0x0
fault code              = supervisor read instruction, page not present
instruction pointer     = 0x20:0x0
stack pointer           = 0x28:0xffffff812aced940
frame pointer           = 0x28:0xffffff812aced9f0
code segment            = base 0x0, limit 0xfffff, type 0x1b
                        = DPL 0, pres 1, long 1, def32 0, gran 1
processor eflags        = interrupt enabled, resume, IOPL = 0
current process         = 1716 (zfs)
trap number             = 12
panic: page fault
cpuid = 1
KDB: stack backtrace:
#0 0xffffffff809889b6 at kdb_backtrace+0x66
#1 0xffffffff8094f2ce at panic+0x1ce
#2 0xffffffff80d038e0 at trap_fatal+0x290
#3 0xffffffff80d03c41 at trap_pfault+0x211
#4 0xffffffff80d041f4 at trap+0x344
#5 0xffffffff80ced4e3 at calltrap+0x8
#6 0xffffffff808367db at devfs_ioctl_f+0x7b
#7 0xffffffff8099b3f6 at kern_ioctl+0x106
#8 0xffffffff8099b63d at sys_ioctl+0xfd
#9 0xffffffff80d0308a at amd64_syscall+0x5ea
#10 0xffffffff80ced7c7 at Xfast_syscall+0xf7
Uptime: 7m36s
Automatic reboot in 15 seconds - press a key on the console to abort
Rebooting...
cpu_reset: Restarting BSP
cpu_reset_proxy: Stopped CPU

NOTE: This is a FreeBSD Xen DomU
% uname -a
FreeBSD c3po.geek.lan 9.1-STABLE FreeBSD 9.1-STABLE #0: Tue Jul 23 17:49:39
UTC 2013     root at c3po:/usr/obj/usr/src/sys/XENHVM  amd64

After booting in single user and removing the flag I was able to boot
normaly and start the jails!

I can use mount_nullfs to see the zfs filesystem inside the jail, but I
still would like to know why I get the panic!

I'm also not sure whats the best way to make zfs datasets visible inside a
jail... should simply use mount_nullfs?

Thanks