Subject: SCSI/VM/?? bug/error?
To: None <amiga@NetBSD.ORG>
From: Greg Oster <oster@skdad.usask.ca>
List: amiga
Date: 01/21/1995 23:47:25
Hello NetBSD fans...

What does the following mean:

sd4(ahsc0:4:0): medium error, info = 273200 (decimal)

I havn't seen a message similar to this for quite some time (pre 1.0,
perhaps?) , but it happened again this evening.  I couldn't reproduce
it before, and I still can't now.  At the same time the error happened
an xterm "disappeared" (It was the xterm disappearing that prompted me
to check for the error, and the times were right).

My guess that it's some sort of a VM or SCSI error, but I don't know
for sure...

Ideas?

----------

The system: 

A3000 68030@25 68882@25 16MB FAST 2MB CHIP
Retina ZII w/ 4MB
Quantum 105 (ADOS only) ID 6
Maxtor 7213S (ADOS/NetBSD) ID4 - swap and root are on this drive.  
	No other NetBSD partitions on this drive.
Quantum PD1225S (NetBSD only) ID 1 - all other NetBSD paritions on
	this drive.
WangTek 5150ES Tape Drive ID 2

Proto 00-04 SCSI Chip  
Superkickstart

----------

Kernel: self compiled from NetBSD 1.0 sources.  Options the same as
the "A3000" options in conf directory, except for:

  added GATEWAY FDESC PANICWAIT DDB ADOSFS LKM sl ppp + Retina support
  changed # of users from 16 to 64
  changed # of pty's from 16 to 128

(still using "config netbsd swap on generic")
----------

"/sbin/dmesg" starts with:

NetBSD 1.0 (AMIGA) #0: Wed Jan  4 21:55:26 CST 1995
    oster@gonzo.usask.ca:/src/sys/arch/amiga/compile/AMIGA
Amiga 3000 (m68030 CPU/MMU m68882 FPU)
real  mem = 16777216 (2048 pages)
avail mem = 14671872 (1791 pages)
using 115 buffers containing 942080 bytes of memory

------- 

The current "uptime" is a little over 17 days, and at the time of the
"error" the load was about 3.40.  Swap used was about 30MB of 90MB (I
played around the other day and was using about 89.4MB out of 90MB -
didn't dare fill it up much more than that for fear of a crash :-/ )

No ADOS partitions were mounted at the time.

Has anyone else seen this error?  Is there anything I can do to help
track this down (e.g. would turning on the SCSI debugging stuff help?
- that and run the machine into the ground in the hopes of
re-producing the error?)?  Could it be a media error?  (Personally, I
don't think it is.)

If you need any more info just let me know...

Thanks... (And keep up the good work...)

Later...

Greg Oster

oster@cs.usask.ca
Department of Computational Science
University of Saskatchewan, Saskatoon, Saskatchewan, CANADA