Regression: Panic on shutdown with CURRENT from March 5, 2007

After upgrading my current 2 days ago I noticed this while doing a 'halt -p'

Uptime: 1h42m2s


Fatal trap 12: page fault while in kernel mode
cpuid = 0;  apic id = 00
fault virtual address		= 0x0
fault code				= supervisor read, page not present
instruction pointer		= 0x20:0xc06dc87e
stack pointer			= 0x28:0xe6d2fb08
frame pointer			= 0x28:0xe6d2fb0c
code segment			= base 0x0, limit 0xfffff, type 0x1b
					= DPL 0, pres 1, def32 1, gran 1
processor eflags		= interrupt enabled, resume, IOPL = 0
current process			= 1603 (halt)
[thread pid 1603 tid 100211 ]
Stopped at	device_shutdown+0x22:	cmpl	$0xc09efb7c,0(%eax)

Running 6.2R and an old CURRENT (some months old) all is fine.
Now, it always does this, easy to replicate, so I can get more info if
instructed to, just have to copy it by hand :|

-- 
Joao Barros