On Thu, 24 Aug 2000, Sean Doran wrote:

# Date: Thu, 24 Aug 2000 23:48:56 +0200 (CEST)
# From: Sean Doran <smd@ebone.net>
# To: smd@ebone.net, tech-kern@netbsd.org
# Subject: Re: lseek(2) & read(2) on raw cd(4) device
# 
# Ah... some of the "problem" was from (dev/scsipi/cd.c, cdstrategy):
# 
#         /*
#          * Do bounds checking, adjust transfer. if error, process.
#          * If end of partition, just return.
#          */
#         if (CDPART(bp->b_dev) != RAW_PART &&
#             bounds_check_with_label(bp, lp,
#             (cd->flags & (CDF_WLABEL|CDF_LABELLING)) != 0) <= 0)
#                 goto done;
# 
# Note the check for whether we are using RAW_PART.  Using the non-raw 
# partitions (/dev/rcd0a, /dev/rcd1a) seems to make pax happy when reading.

um, /dev/r?d[0-9]*[a-h] are raw partitions.

/dev/?d[0-9]*[a-h] are block partitions.

Or do I mis-understand your terminology?

# 	Sean.

				--*greywolf;
--
BSD: The devil's advocate.