> 
> All process running on cpu1 cease running. No more processes can
> be scheduled on it. None of the processes running on it can be
> killed with 'kill -9'. All activity associated with that cpu is
> dead.

From experiences with other systems I'd suspect the cpu is an
a screaming loop waiting for a lock.

Can be hard to find out where though.

With a few diagnostics you ought to be able to determine
the active process and whether it is in user or kernel.
The next bit involves parsing the cpu stack to sort out a
likely PC value - quite tricky since you have to do it
backwards!

OTOH maybe the inter-cpu call (used for TLB shootdown etc)
could be persuaded to dump some state in global memory.

	David

-- 
David Laight: david@l8s.co.uk