can you "call simple_lock_dump" from ddb?  your stack trace
seems to show both processes handling syscalls?  i guess cpu1
is currently spinning waiting for the kernel lock, from the
track...


.mrg.