Subject: Re: SCSI DMA kernel with "glass tty" support.
To: Anders Magnusson <ragge@ludd.luth.se>
From: NetBSD Mailing list <netbsd@mrynet.com>
List: port-vax
Date: 10/26/1999 15:39:50
The new DMASCSI-CRT kernel works great.   

However... :)

I can access partition c just fine on my newly labelled
SCSI drive.  I can't access any other partition.
I can dd, newfs, etc... on /dev/rsd0c.   Accessing any
other partition returns:

# newfs /dev/rsd0a
newfs: /dev/rsd0a: open: Device not configured
# dd if=/dev/rsd0a
dd: /dev/rsd0a: Device not configured
# cat /dev/rsd0a
cat: /dev/rsd0a: Device not configured

Disklabel (both with and without -r) reports:
# disklabel sd0
# /dev/rsd0c:
type: SCSI
disk: hp3725
label: 
flags:
bytes/sector: 512
sectors/track: 125
tracks/cylinder: 9
sectors/cylinder: 1125
cylinders: 3703
total sectors: 4165875
rpm: 3600
interleave: 1
trackskew: 0
cylinderskew: 0
headswitch: 0           # milliseconds
track-to-track seek: 0  # milliseconds
drivedata: 0 

8 partitions:
#        size   offset     fstype   [fsize bsize   cpg]
  a:    15884        0     unused     1024  8192         # (Cyl.    0 - 14*)
  b:    66880    16875       swap                        # (Cyl.   15 - 74*)
  c:  4165875        0     unused     1024  8192         # (Cyl.    0 - 3702)
  d:    15884   375750     unused      512  4096         # (Cyl.  334 - 348*)
  e:   307200   392625     unused     1024  8192         # (Cyl.  349 - 622*)
  f:  3465000   700875     unused     1024  8192         # (Cyl.  623 - 3702)
  g:  3790125   375750     unused     1024  8192         # (Cyl.  334 - 3702)
  h:   291346    84375     unused     1024  8192         # (Cyl.   75 - 333*)

This is indeed after rebooting, so any desync from a fresh disklabel
would not be an issue.  I've redone the /dev entries, and I am using
disklabel/newfs compiled from a sup(1) from yesterday.


And, regarding the le0 timeouts that have been mentioned, I am seeing 
this timeout constantly.  This machine (while I am setting up the SCSI
drive) has root over NFS, and I'm also telnetting.  When I say, "constantly,"
I mean every few minutes.  This timeout is definately an issue needing
resolution.  I don't see the le0 timeout at all when using kernels from
a month ago or so in this same setup of netbooting and telnetting.

Regards,
			-skots