Well, I've looked at this some more ... and looked at Rob's suggestion 
re: dk_ndrive.

I have a few things to say about this.  For one, dk_ndrive is apparently 
here only so that user-land (specifically iostat and vmstat) can find out 
how big the disk statistics array is.  It's almost worthy of a sysctl 
(a'la kern.maxpartitions).  It is, therefore, not the appropriate 
solution to this problem.  For an explanation why, grep for it in 
sys/kern_clock.c.

Secondly, disk statistics gathering is _BROKEN_!  There are a number of 
reasons for this, one of them being the MI SCSI code doesn't have glue it 
in for dk stats, and the dkdevice structure doesn't have a way of storing 
an index.

The correct solution is, of course, to fix the dk stats system.  In the 
mean time, I'm going to #ifdef out the stats glue in ccd.c (/* XXX */).

--------------------------------------------------------------------------
Jason R. Thorpe                                       thorpej@nas.nasa.gov
NASA Ames Research Center                               Home: 408.866.1912
NAS: M/S 258-6                                          Work: 415.604.0935
Moffett Field, CA 94035                                Pager: 415.428.6939