I built a new kernel from 11/17 sources yesterday, to replace
the 11/3 kernel I was using before. If I've been reading the messages
properly, this is my first kernel with the new pmap code in it. It
seemed to work, making it most of the way through boot, but right
after I got the log message of "kernel security level changing from 0
to 1", it paniced:

panic: pmap_to_pvflags: idx=0x7f800

DDB backtrace:

_Debugger(0,f217f1c,e0a1e00,e09fe90,7f800) + 6
_panic(e09fe90,7f800,0,e72b380,f217f1c) + 40
_pmap_clear_modify(ff000010) + 26
_vm_fault(e5d8500,2120000,3,0,eeeeeeee) + 72
__pmap_fault(e6d8500,2120000,3) + 46
_trap(8,101,2120000) + 448
faultstkadj() + 0

gdb backtrace:
(gdb) where
#0  cpu_reboot (howto=256, user_boot_string=0xe72ba80 "\016rĄ")
    at ../../../../arch/sun3/sun3/machdep.c:558
#1  0xe09f698 in cpu_reboot (howto=256, user_boot_string=0x0)
    at ../../../../arch/sun3/sun3/machdep.c:555
#2  0xe01dc04 in panic (fmt=0xe09fe90 "pmap:pa_to_pvflags: idx=0x%x")
    at ../../../../kern/subr_prf.c:150
#3  0xe0a1e00 in pmap_clear_modify (pa=4278190096)
    at ../../../../arch/sun3/sun3/pmap.c:479
#4  0xe08c9d2 in vm_fault (map=0xe5b1f00, vaddr=34734080, fault_type=3, 
    change_wiring=0) at ../../../../vm/vm_fault.c:354
#5  0xe0a1cf0 in _pmap_fault (map=0xe5b1f00, va=34734080, ftype=3)
    at ../../../../arch/sun3/sun3/pmap.c:2539
#6  0xe0a3ed0 in trap (type=8, code=257, v=34734080, tf={tf_regs = {34, 0, 
        -286331154, 1, 899, 0, 0, -1, 756268, 34734084, 34734080, 4, 1105444, 
        234880340, 234880148, 234880088}, tf_pad = 0, tf_stackadj = 0, 
      tf_sr = 8, tf_pc = 420994, tf_format = 10, tf_vector = 8})
    at ../../../../arch/sun3/sun3/trap.c:501


	I still have the corefile, if there's more useful debugging I
can do. Ideas? It's clearly from a DIAGNOSTIC check, so there's a real
problem here somewhere.

	- Nathan