strange LOR/lockup

I'm a bit confused by the following lock order reversal. 
My box started by locking in 7.1-PRERELEASE. However, with debugging
and WITNESS, I saw the LOR, but the console still became unresponsive.

I upgraded to CURRENT #0: Thu Nov 27 17:51:41 PST 2008, and the
problem still persists.  I'm not sure what exactly is happening, or
how to debug it further.

Actually, by triggering some sort of console message (removing a usb
device, triggering an acpi error with my power switch) I can see the
keyboard input and console output, but it really doesn't work very
well. Essentially, at that point a hard powerdown is all I can do.


kernel: lock order reversal:
kernel: 1st 0xffffff000c7f3550 rtentry (rtentry) _at_ /mnt/src/sys/net/route.c:333
kernel: 2nd 0xffffff0004e82af8 radix node head (radix node head) _at_ /mnt/src/sys/net/route.c:887
kernel: KDB: stack backtrace:
kernel: db_trace_self_wrapper() at db_trace_self_wrapper+0x2a
kernel: _witness_debugger() at _witness_debugger+0x2e
kernel: witness_checkorder() at witness_checkorder+0x81e
kernel: _mtx_lock_flags() at _mtx_lock_flags+0x78
kernel: rtrequest1_fib() at rtrequest1_fib+0x8f
kernel: rtredirect_fib() at rtredirect_fib+0x28e
kernel: icmp_input() at icmp_input+0x3b1
kernel: ip_input() at ip_input+0xc0
kernel: ether_demux() at ether_demux+0x1ed
kernel: ether_input() at ether_input+0x1d5
kernel: em_rxeof() at em_rxeof+0x200
kernel: em_handle_rxtx() at em_handle_rxtx+0x4b
kernel: taskqueue_run() at taskqueue_run+0x96
kernel: taskqueue_thread_loop() at taskqueue_thread_loop+0x55
kernel: fork_exit() at fork_exit+0x12a
kernel: fork_trampoline() at fork_trampoline+0xe
kernel: --- trap 0, rip = 0, rsp = 0xfffffffe400bcd40, rbp = 0 ---