Assuming you can type at all at the "db>" prompt, type "trace" to get 
a stack backtrace (i.e. where the kernel was executing) so that a 
better notion of what the kernel was doing when it died can be had...

	Erik