crash in natd again

This is on an amd64 machine.  I though that setting

dumpdev="AUTO"

in /etc/rc.conf was sufficient to get a crash dump, but it didn't work.
I'm adding

dumpon_enable="YES"
savecore_enable="YES"

now.  Please let me know if there is anything else I need to set.

The information I can get is:

Nov 22 13:47:27 server kernel: Fatal trap 12: page fault while in kernel mode
Nov 22 13:47:27 server kernel: fault virtual address    = 0x18
Nov 22 13:47:27 server kernel: fault code               = supervisor read, page
not present
Nov 22 13:47:27 server kernel: instruction pointer      = 0x8:0xffffffff80357980Nov 22 13:47:27 server kernel: stack pointer            = 0x10:0xffffffffb41c5620
Nov 22 13:47:27 server kernel: frame pointer            = 0x10:0xffffffffb41c5680
Nov 22 13:47:27 server kernel: code segment             = base 0x0, limit 0xfffff, type 0x1b
Nov 22 13:47:27 server kernel: = DPL 0, pres 1, long 1, def32 0, gran 1
Nov 22 13:47:27 server kernel: processor eflags = interrupt enabled, resume, IOPL = 0
Nov 22 13:47:27 server kernel: current process          = 215 (natd)
Nov 22 13:47:27 server kernel: trap number              = 12
Nov 22 13:47:27 server kernel: panic: page fault
Nov 22 13:47:27 server kernel: KDB: enter: panic

uname is

FreeBSD server.mcneil.com 6.0-CURRENT FreeBSD 6.0-CURRENT #5: Sun Nov 21
00:33:27 PST 2004     root_at_server.mcneil.com:/usr/obj/usr/src/sys/AMD64
amd64

(gdb) l *0xffffffff80357980
0xffffffff80357980 is in m_copym (/usr/src/sys/kern/uipc_mbuf.c:373).
368             MBUF_CHECKSLEEP(wait);
369             if (off == 0 && m->m_flags & M_PKTHDR)
370                     copyhdr = 1;
371             while (off > 0) {
372                     KASSERT(m != NULL, ("m_copym, offset > size of
mbuf chain"));
373                     if (off < m->m_len)
374                             break;
375                     off -= m->m_len;
376                     m = m->m_next;
377             }