netbsd-6 and large memory systems

[John Klos <john%ziaspace.com@localhost>]

Hi,

Although I don't have any real systems with more than 128 megs of memory, 
I was playing with simh and noticed that while netbsd-5 has no problem 
with 512 megs, a netbsd-6 kernel panics:

boot device: ra0
root on ra0a dumps on ra0b
root file system type: ffs
Thu Apr 12 14:02:14 PDT 2012
r0=9fe652a0 r1=8027f240 r2=9fe652a0 r3=00000000 r4=80c10160 r5=9fdaa801 
r6=9fd650b4 r7=00000010
r8=9fdbb420 r9=9fd650b4 r10=9ff65000 r11=800e7070
ap=a717aadc fp=a717aac8 sp=7fffe378 pc=800e71e2
panic: SEGV in kernel mode: pc 0x800e71e2 addr 0xbedababe
Stopped in pid 5.1 (rcorder) at netbsd:printf_nolog:    function 
"printf_nolog()
", entry-mask 0x0
                remqhi  *0xedf8ef00(r11), *0x88c4fa92
db> bt
panic: SEGV in kernel mode: pc 0x800e71e2 addr 0xbedababe
Stack traceback :
0xa717a974: printf_nolog+0x0(0x80261f37,0x800e71e2,0xbedababe)
0xa717a9a4: trap+0x721(0xa717aa78)
0xa717aa78: trap type=0x8c code=0xbedababe pc=0x800e71e2 psl=0xc00004
0xa717aa44: _do_cas+0xe(0x8027f240)
0xa717aac8: vremfree+0x36(0x9fd650b4)
0xa717aae8: vget+0x1b(0x9fd650b4,0x10)
0xa717ab10: cache_lookup+0x15d(0x9fdbb420,0xa717ac70,0xa717aea8)
0xa717ab40: ufs_lookup+0x8a(0xa717ac2c)
0xa717abf8: VOP_LOOKUP+0x41(0x9fdbb420,0xa717ac70,0xa717aea8)
0xa717ac40: lookup_once+0xf9(0xa717ad80,0x9fdbb420,0xa717ad38,0xa717ad34)
0xa717ac78: namei_tryemulroot+0x18a(0xa717ad80,0,0,0)
0xa717ad40: namei+0x2e(0xa717ae84)
0xa717ad9c: vn_open+0x7c(0xa717ae84,0x1,0x1a4)
0xa717ae48: do_open+0x7a(0x9fe652a0,0x9ffb85a0,0,0x1b6,0xa717af14)
0xa717aed0: sys_open+0x3f(0x9fe652a0,0xa717af4c,0xa717af74)
0xa717af20: syscall+0x93(0xa717afb4)
db>


simh only allows powers of 2 memory sizes, so I can only say that netbsd-6 
boots with 256 megs but not with 512.

John