Problems booting OSR5.0.2 on notebook w/ EIDE 2GB drive

Problems booting OSR5.0.2 on notebook w/ EIDE 2GB drive

Post by Matthew Thurmaie » Tue, 17 Sep 1996 04:00:00



I have installed SCO OSR5 on a HyperData HyperBook (which resolves to a:
* 133 MHz Pentium
* 32Mb RAM
* dual EIDE controllers, disk on one, ATAPI CD-ROM on secondary
* Touch-pad mouse (works like Lowres keyboard mouse)
* Sound-blaster compatible sound
* PC-Card slot w/ 3com 589 PCMCIA networking card)

After some work and much frustration, I found out that I could get
the system to boot off floppy and see the root hard disk and ATAPI
CDROM with the following:
        biosgeom=(4200,16,63) defbootstr disable=dptr prompt="watch"
(I put the prompt in so that during the long floppy load, I wouldn't
 miss the kernel configuration line).

The OS installed fine.  During badtrk, some sectors were marked bad in
the area of cyl 100.  No big deal.

When I go to re-boot after the install, the system:
1) hits the disk 5 times (you can hear it)
2) prints the message "IO ERROR"

The manual page for boot(HW) says that the "IO ERROR" message comes
from the master boot program (which resides in the master boot block,
i.e. block 0) tries to read in the boot program on the active (UNIX)
partition. (parens my own).  So, I tried:
1) booting to the boot prompt off floppy and typing:
        biosgeom=(4200,16,63) hd(40)unix
   which results in the message:
        Error on Fixed Disk (minor 15), blkno=2, cmd=0002, status=0004
        Sectpr = 67, Cylinder/head = 0/1
        Cannot open device: hd(40)
2) booting w/ the kernel off the floppy and rooting off the hard disk.
   I do this with:
        biosgeom=(4200,16,63) fd(64)unix.Z root=hd(42) swap=hd(41)
   This works fine.  But I don't _want_ to have to boot off floppy every
   time and I shouldn't have to either.
3) running dparam and re-writing the disk geometry info out to disk
4) copying /etc/masterboot to /dev/hd00
5) re-making the master boot code with:
        cp /etc/hdboot0 /dev/hd0a
        dd if=/etc/hdboot1 of=/dev/hd0a bs=1k seek=1

One thing I've noticed along the way... the biosgeom flag appears to
have a bug in that the disk geometry reported by the "%disk" line at
kernel boot is one more than that provided in the biosgeom string.  So,
I have tried this all a second time with biosgeom set to (4199,16,63)
and that didn't work either.

OK, so what am I doing wrong or not doing?

Thanks,
Matthew
--
************************************************************************
* Matthew J. Thurmaier - President  * We teach Novell/Sco/Sun Unix and *
* The Computer Classroom, Inc.      * will soon be offering Microsoft  *
* 437 S. Yellowstone, Dr., Ste 112  * WIN95 & NT courses.  Remember    *
* Madison, WI 53719, 800-603-8988   * Education is an investment, not  *
* http://www.compclass.com          * an expense.                      *
************************************************************************
* The Computer Classroom, Inc. has been selected as SCO's AEC of the   *
*             year for 1996, the fourth year of the past five!         *
************************************************************************

 
 
 

1. tar, HP DAT drive, OSR5.0.2 - help?

I was hoping someone could help me with this problem.  My server has 2
Pentium 150s, an Adaptec HBA (the 7770 chip, using the arad driver), a
Mylex DAC960 RAID controller (using Mylex's driver for SCO Unix) with 5
2GB drives, an HP DAT drive, CDROM, and a Wangtek cartridge tape drive.  
The last couple of times I've tried to migrate some files from my other
server (running 3.2 v4.2) by using a tar volume on a DAT tape, I've
gotten these error messages:

CPU2: WARNING: arad: adapter 0 Error: Target Bus Phase Sequence Error
tar: tape read error
read: I/O error
NOTICE: Sdsk: Unrecoverable error writing SCSI disk 0 dev 1/42 (ha=0 id=0
   lun=0) block=82074
   Unit attention:Drive or bus reset

This happens after the tape has been extracting for a while, and it kicks
me out to a prompt.  However, the system then hangs no matter what I try
to do.  I've heard that that may be some issues between SCO and the HP
DAT drives, could this be related? And, if so, what is the fix?
   Fortunately, this server isn't in use yet, but I need the problem
solved.  I do recall that when I installed OSR5.0.2. it balked a little
when I tried to link the arad driver, saying that an "aradinit line
already exists" in a table ( I know, I know, I should have written it
down exactly) somewhere, and asking me if I wished to add to the end of
that file, replace the line, or do nothing, I chose to replace.  Later on
in the install process, it again complained about the arad driver, this
time saying it was missing, but the install process continued, and seemed
to complete successfully.
  The system had been working well up until this problem with the tape
drive, I hope someone can help me.  Sorry for the length of the post, but
I wanted to try and give everyone enough info to make informed judgments.

TIA,
Mark

2. Should sysinstall display ERRATA.TXT?

3. OSR5.0.2 won't boot after CPU replacement

4. Help with bind configuration

5. Win95-OSR5.0.2 Dual Boot Possible?

6. hostname/IP spoofing

7. HTFS problem on SCO OSR5.0.2

8. impsec.org ?? ?

9. SCO OSR5.0.2 - X11 License Manager problem

10. Problem with SCO OSR5.0.2 / AFPS/ Windows 95

11. OSR5.0.2 swap problem

12. Lone-Tar Backup Errors ==> OSR5.0.2 Problems