Please help with A3000 SCSI disk problem

[Larry <lstone%gte.net@localhost>]

I hope this is the correct list for posting this question. It seems to
me to be very specific to the Amiga. I have been following this list for
several weeks now, and have dug into the archives, but have not seen any
mention of this problem.

I am trying to install NetBSD 1.2.1 on my A3000. It is a pretty stock
machine (25 MHz, 2 MB chip, 16 MB fast, all on the motherboard), with
the following cards in the expansion slots:
        A4066 Ethernet
        Emplant (SCSI chain is currently activated)
        MultiFaceCard III
I have the proto version of the SCSI chip, and I have three hard drives
installed:
        SCSI Id 0       Seagate ST1096N (80 MB)
        SCSI Id 1       Maxtor 7345 (345 MB)
        SCSI Id 2       Compaq 0662S12 (SCSI2, 1 GB)
I want to install NetBSD on the Compaq drive.

I go through the installation process as described in the INSTALL
document, and everything goes well until I try to actually install the
base packages. I partitioned the drive, copied the installation miniroot
into the swap partition, booted the kernel from AmigaDOS and mounted my
AmigaDOS disk on /mnt/ados. When I run the command to gunzip/tar the
installation files, it copies about 40-60k to the root partition, and
then I start getting SCSI driver error messages, which scroll by until I
^C the process. The message is:

sd2(ahsc0:2:0): command aborted, data = 00 00 01 61  00 00 00 00\
                                        00 00 00 00  00 00 00 00\
                                        00 00 00 00  0a 00 05 7a
I formatted the hex values to make them easier to interpret, for someone
who knows what they mean. The first 4 numbers are always the same, for
every test I ran. The middle numbers are almost always all zeroes. The
last 4 numbers vary from test to test, but are usually consistent within
a test (sometimes the last number changes after 5 or 10 messages).

The installation file suggests booting with synchronous transfer
disabled if the bootup process hangs. Well, I'm not hanging, but I gave
it a try anyway. The bootup process hangs trying to access disk 2 (the
problem child) when I boot with the command given in the file:
        loadbsd -I ff -b netbsd

After playing with the commands for awhile (I dug through the script and
ran the command directly, then I just started trying to copy the files),
I was unable to get anything onto the disk. I then tried echoing a short
string to a file on the root partition. I then cat'd this file with
itself into a new file, and kept repeating this, doubling the file size
each time. I was able to build the file up to 1.5 MB this way, with no
errors reported. However, a straight copy of a much smaller file (one of
the base installation files) failed every time.

So, now I'm really confused. I switched back to AmigaDOS, reset the file
system type for one of the partitions on sd2, formatted it in ADOS, and
copied megabytes to it with no problem at all. I then changed the
Seagate drive to be a NetBSD "other" partition, and newfs'd it from the
NetBSD side, and was again able to copy megabytes to it with no errors
reported. So this is something specific to the Compaq drive, and to the
NetBSD driver.

The only things I can think of are:

1. My SCSI termination isn't right. All 3 drives are internal, and I
have nothing connected to the external SCSI port. The drive order from
the motherboard is: Compaq, Maxtor, Seagate(terminated). I don't have
specs on the Compaq drive, and it doesn't have the resistor pack holes
I'm used to. I don't think it's terminated, but I'm not positive. If it
were terminated, wouldn't the other two drives be invisible?

2. The prototype chip doesn't work with the NetBSD 1.2.1 SCSI drivers.

3. My system can't handle the SCSI-2 drive (perhaps because of the
prototype chip?).

4. I need to try some special parameters when newfs'ing the drive.

5. Somehow, the Emplant SCSI device is messing with my transfer (doesn't
seem very likely to me).

Can someone out there suggest what I should try next? I have ordered a
replacement SCSI chip, figuring that (a) it's not very expensive and (b)
even if it doesn't help, it shouldn't hurt.

I'm going to include the bootup messages, in case they will be of any
help. I have the output of a disklabel command run on each of the
drives, and would be happy to send them if they would help. I'll go
ahead and include the disklabel for sd2, since that's the drive in
question.

Thanks in advance to anyone who can help,

        Larry W. Stone
        lstone%gte.net@localhost

-----------------------------------------------------------------------
Here are the bootup messages. I copied them by hand, so if something
looks really weird it's probably because I made a mistake. I'll be happy
to reboot the machine and check anything that looks bizarre.

Amiga 3000 (m68030 CPU/MMU m68882 FPU)
real  mem = 16777216 (2048 pages)
avail mem = 14221312 (1736 pages)
using 155 buffers containing 942080 bytes of memory
memory segment 0 at 7000000 size 1000000
memory segment 1 at 0 size 200000
mainbus0 (root)
clock0 at mainbus0: CIA B system hz 100 hardware hz 715909
ser0 at mainbus0: input fifo 512 output fifo 32
par0 at mainbus0:
kbd0 at mainbus0: CIA A type Amiga
ms0 at mainbus0
ms1 at mainbus0
grfcc0 at mainbus0
grf0 at grfcc0: width 640 height 400 colors 4
ite0 at grf0: rows 50 cols 79 repeat at (30/100)s next at (10/100)s has
keyboard
fdc0 at mainbus0: dmabuf pa 0x1e3030: dmabuf ka 0x2ddf030
fd0 at fdc0 unit 0: 3.5dd 80 cyl, 2 head, 11 sec (9 sec), 512 bytes/sec
zbus0 at mainbus0
es0 at zbus0 rom 0xe90000 man/pro 1053/10: address 00:00:9f:fe:00:a3
empsc0: sci id 7
scsibus0 at empsc0
mfc0 at zbus0 rom 0xcb0000 man/pro 2092/18
mfcs0 at mfc0: input fifo 1024 output fifo 128
mfcs1 at mfc0: input fifo 1024 output fifo 128
ahsc0 at mainbus0
scsibus1 at ahsc0
sd0 at scsibus1 targ 0 lun 0: <SEAGATE, ST1096N, \003\024\012> SCSI1
0/direct fixed
sd0: 80 MB, 906 cyl, 7 head, 25 sec, 512 bytes/sec
ahsc0: target 1 now synchronous, period = 208ns, offset = 12.
sd1 at scsibus1 targ 1 lun 0: <MAXTOR, 7345-SCSI, 0960> SCSI1 0/direct
fixed
sd1: 329 MB, 2220 cyl, 4 head, 76 sec, 512 bytes/sec
ahsc0: target 2 now synchronous, period = 208ns, offset = 12.
sd2 at scsibus1 targ 2 lun 0: <COMPAQ, 0662S12, 3C31> SCSI2 0/direct
fixed
sd2: 1001 MB, 4119 cyl, 5 head, 99 sec, 512 bytes/sec
... then it asks for the root partition

-----------------------------------------------------------------------
This is the output of the disklabel command for sd2. Note that this is
after I changed the last partition to AmigaDOS format. That partition
was originally set up for 4.2BSD, just like the "d" partition.

type: SCSI
disk: SCSI disk
label: BSD_SWAP
flags:
bytes/sector: 512
sectors/track: 525
tracks/cylinder: 1
sectors/cylinder: 525
cylinders: 3906
total sectors: 2050650
rpm: 3600
interleave: 1
trackskew: 0
cylinderskew: 0
headswitch: 0           # milliseconds
track-to-track seek: 0  # milliseconds
drivedata: 0 

5 partitions:
#        size   offset    fstype   [fsize bsize   cpg]
  a:    83475     1050    4.2BSD     1024  8192    16   # (Cyl.    2 -
160)
  b:    98700    84525      swap                        # (Cyl.  161 -
348)
  c:  2050650        0    unused        0     0         # (Cyl.    0 -
3905)
  d:   511875   183225    4.2BSD     1024  8192    16   # (Cyl.  349 -
1323)
  e:  1355550   695100      ADOS                        # (Cyl. 1324 -
3905)
disklabel: boot block size 0
disklabel: super block size 0