sjr@mimsy.UUCP (Stephen J. Roznowski) (10/27/87)
I apologize for this being so long, but I'm trying to be precise as to
what my problem is.
I'm currently having problems booting 4.2BSD on a VAX 11/730 with the
following configuration:
2 Meg Memory
RL02/R80 on IDC controller
DMF32
TU80 Tape drive
Excelan Ethernet card
Version 54 of the console microcode (BE-T173C-DE console tape)
As far as DEC is concerned the machine is ok. All their diagnostics
pass and they are able to bring VMS up on the machine.
The problems started when DEC replaced the HDA in the R80 (well, actually
the problems started when the HDA failed .....) Since the HDA has been
replaced, I've been unable to bring 4.2 up on the hardware.
I've been able to recreate all the problems below with a minimal version
of the system running (i.e. no DMF32 and no Excelan).
The HDA was formatted with EVRGA (RM80/IDC R80 FORMATTER) and found 0
bad sectors and 1 skip sector (Logical Blk 132565, Cyl 305, Trk 6, Sec 9).
-----
I have a bootable copy of 4.2BSD on the RL02 and when I try to create
partition on the R80 (drive #1) I get:
newfs -n -v /dev/rrb1a rb80
/etc/mkfs /dev/rrb1a 15884 31 14 8192 1024 16 10 60 2048
rb1a: hard error sn15883 \
csr=41095c1<R80,ECSO,ERR,DLT,OPI,DSO,CRDY,IE,DRDY> ds=0
write error: 15883
wtfs: I/O error
A similiar thing happens on the other partitions:
/dev/rrb1h errors on sn110142
/dev/rrb1g errors on sn82079
The number is always one less then the number reported by newfs. (after
the /dev/rrb??)
-----
I have tried using dd to read the raw partition with the following
results:
dd if=/dev/rrb1c of=/dev/null bs=1b
rb1c: hard error sn434 \
csr=41095c1<R80,ECSO,ERR,DLT,OPI,DSO,CRDY,IE,DRDY> ds=0
read: I/O error
0+0 records in
0+0 records out
dd if=/dev/rrb1c of=/dev/null bs=1b skip=437
rb1c: hard error sn434 \
csr=41095c1<R80,ECSO,ERR,DLT,OPI,DSO,CRDY,IE,DRDY> ds=0
rb1c: hard error sn435 csr=41095c1<same as above> ds=0
rb1c: hard error sn436 csr=41095c1<same as above> ds=0
rb1c: hard error sn437 csr=41095c1<same as above> ds=0
read: I/O error
0+0 records in
0+0 records out
-----
I have tried to create an extremely small file system:
/etc/mkfs /dev/rrb1a 434 31 14 4096 512 4 10 60 256
/dev/rrb1a: 434 sectors in 1 cylinders of 14 tracks, 31 sectors
0.2Mb in 1 cyl groups (4 c/g, 0.89Mb/g, 64 i/g)
super-block backups (for fsck -b#) at:
32,
cg 0: bad magic number
cg 0: bad magic number
I at least understand the cg 0 errors, mkfs is unable to build
a file system on 1 cylinder. At least it looks like mkfs is working.
(as opposed to just erroring on the last sector)
-----
Finally, I have tried to use standalone copy to boot of a 4.2 tape.
>>>L DD1:COPY
>>>S 2
From: ts(0,1)
To: rb(1,1)
idc error: (cyl,trk,sec)=(410,0,0) \
csr=3958b<ATN1,ATN0,ERR,DLT,OPI,DSO,CRDY,F2,F0,DRDY>
idc recovered by retry
idc error: (cyl,trk,sec)=(410,1,0) csr=3958b<same as above>
idc recovered by retry
idc error: (cyl,trk,sec)=(411,0,0) csr=3958b<same as above>
idc recovered by retry
idc error: (cyl,trk,sec)=(411,1,0) csr=3958b<same as above>
idc recovered by retry
idc error: (cyl,trk,sec)=(412,0,0) csr=3958b<same as above>
idc recovered by retry
idc error: (cyl,trk,sec)=(412,1,0) csr=3958b<same as above>
idc recovered by retry
idc error: (cyl,trk,sec)=(413,0,0) csr=3958b<same as above>
idc recovered by retry
?02 PC=00001B60
>>>
After the last idc error, the system halts.
However doing
>>>L DD1:COPY
>>>S 2
From: ts(0,1)
To: rb(0,1) <-- the RL02
idc error:(cyl,trk,sec)=(512,0,0) \
csr=3948b<ATN1,ATN0,ERR,DLT,OPI,CRDY,F2,F0,DRDY>
idc: recovered by retry
Copy completed: 205 records copied
From:
seems to work.
-----
I'm having DEC bring me a replacement IDC controller to see if swapping
it will fix the problem (however, I'm really grasping at straws...)
============
Can anyone help me out?? What am I doing wrong? What else can I try
to do?
Any help would be GREATLY appreciated. (Perhaps if I can't find a
solution I could find volunteers to help me throw the 730 off the
roof :-)
Please respond by Email.
Stephen Roznowski (sjr@mimsy.umd.edu)
--
Stephen J. Roznowski sjr@mimsy.umd.edu