In article <031016234248.M0119406@mirage.ceres.dti.ne.jp>
I wrote:

> This is complete without problem, and after updating all binaries
> on NFS root, build.sh on whole NFS root is also done without errors.

Ugh, right after this post, my O2 kernel died during halt:

---
# halt
Oct 16 14:43:09 halt: halted by root
syncing disks... done
trap: TLB miss (load or instr. fetch) in kernel mode
status=0xff03, cause=0x8, epc=0x8016cb2c, vaddr=0x150
pid=25963 cmd=halt usp=0x7fffec18 ksp=0xc625fb70
Stopped in pid 25963.1 (halt) at        netbsd:nfs_removerpc+0x60:      lw      a0,336(v1)
db> tr
nfs_removerpc+60 (815f4540,c00b680c,803d82e0,c0071400) ra 8016cabc sz 112
nfs_removeit+24 (815f4540,c00b680c,803d82e0,c0071400) ra 8013bb8c sz 32
nfs_inactive+114 (815f4540,c00b680c,803d82e0,c0071400) ra 8020cfdc sz 72
vclean+1f4 (815f4540,c00b680c,803d82e0,c0071400) ra 8020d274 sz 88
vgonel+58 (815f4540,c00b680c,803d82e0,c0071400) ra 8020cd78 sz 32
vflush+f4 (815f4540,c00b680c,803d82e0,c0071400) ra 80162904 sz 48
nfs_unmount+d8 (815f4540,c00b680c,803d82e0,c0071400) ra 80210048 sz 56
dounmount+140 (815f4540,c00b680c,803d82e0,c0071400) ra 8020e754 sz 56
vfs_unmountall+90 (815f4540,c00b680c,803d82e0,c0071400) ra 8020ea44 sz 40
vfs_shutdown+28c (815f4540,c00b680c,803d82e0,c0071400) ra 80294008 sz 64
cpu_reboot+98 (815f4540,c00b680c,803d82e0,c0071400) ra 801dce4c sz 24
sys_reboot+6c (815f4540,c00b680c,803d82e0,c0071400) ra 8026fc20 sz 152
syscall_plain+218 (815f4540,c00b680c,803d82e0,c0071400) ra 80268b74 sz 80
mips3_SystemCall+b8 (815f4540,c00b680c,803d82e0,c0071400) ra 30077270 sz 0
PC 0x30077270: not in kernel space
0+30077270 (815f4540,c00b680c,803d82e0,c0071400) ra 0 sz 0
User-level: pid 25963.1
db> 
---
Izumi Tsutsui
tsutsui@ceres.dti.ne.jp