I built a new kernel last night, with a patch that I got from 
Chris G. Demetriou. I still get problems when I try to access my sd1
disk (a Wren VI, I think; geometry is sectors/track, 15
tracks/cylinder, 1632 cylinders). I'm able to newfs it just fine, but
when I start to copy my sd0 information across (using tar), the system
panics with "ffs_clusteralloc: lost block". When I'm working solely
from sd0 (a 208 MByte Conner), everything seems just fine, so I'm
starting to suspect that the Wren and the pc532 are not at the moment 
on speaking terms. Could it be a timing change in the SCSI code,
perhaps? 

	BTW: I picked up the newest tarballs, and I could compile the
kernel after a small change to arch/pc532/pc532/process_machdep.c
(changed u_int to caddr_t in a function prototype. I think I saw this
file mentioned in one of the "update" mails last week - was there any
other changes?)

	//Raymond.


Raymond Wiker / etorwi@hisoy.etn.ericsson.se / ETO.ETORWI
C/o Ericsson A/S, Boks 44, 4817 HIS, NORWAY / Tel +47 370 51482

------------------------------------------------------------------------------