[comp.unix.microport] 386 UNIX and FDISK

mike@cimcor.mn.org (Michael Grenier) (12/13/88)

Does anyone know if Microport 386 or other versions of AT&T UNIX
for the 386 handle disk drives larger than 1024 tracks? I currently
have Microport 286 version 2.4 and the fdisk command doesn't
support my drive which has 1226 tracks.

	-Mike Grenier
        mike@cimcor.mn.org
	uunet!rosevax!cimcor!mike
	{umn-cs, rutgers, amdahl}!bungia!cimcor!mike

rmarks@KSP.Unisys.COM (Richard Marks) (12/13/88)

In article <618@cimcor.mn.org> mike@cimcor.mn.org (Michael Grenier) writes:
>Does anyone know if Microport 386 or other versions of AT&T UNIX
>for the 386 handle disk drives larger than 1024 tracks? I currently
>have Microport 286 version 2.4 and the fdisk command doesn't
>support my drive which has 1226 tracks.

Hmmm . . . the basic PC disk IO interrupt 13 only has 10 bits for the
track number.

Richard Marks
rmarks@KSP.unisys.COM

phile@lgnp1.LS.COM (Phil Eschallier) (12/14/88)

In article <618@cimcor.mn.org>, mike@cimcor.mn.org (Michael Grenier) writes:
> Does anyone know if Microport 386 or other versions of AT&T UNIX
> for the 386 handle disk drives larger than 1024 tracks? I currently
> have Microport 286 version 2.4 and the fdisk command doesn't
> support my drive which has 1226 tracks.


to the best of my knowledge, AT style machine do not support disk with
more than 1024 tracks.  i have read, however, that you can purchase a
special ROM BIOS, and once you do this, you should be able to type the
drive parameters in via fdisk and all should be well!

i am not sure of any companies that sell such a BIOS, may you could call
the Phoenix folks and ask --

-- anyone know of companies that sell such a monster for the 286 or 386
   machines ??

phil eschallier

dave@pmafire.UUCP (Dave Remien) (12/15/88)

In article <618@cimcor.mn.org> mike@cimcor.mn.org (Michael Grenier) writes:
>Does anyone know if Microport 386 or other versions of AT&T UNIX
>for the 386 handle disk drives larger than 1024 tracks? I currently
>have Microport 286 version 2.4 and the fdisk command doesn't
>support my drive which has 1226 tracks.

My partitions file for a 330 Mb CDC drive on a WD 1007 controller
follows:

disk0:
    heads = 15, cyls = 1222, sectors = 35, bpsec = 512,
    vtocsec = 35, altsec = 36, boot = "/etc/boot", device = "/dev/rdsk/0s0"

(remainder deleted).

All the disk is available, so I'd say that you can get > 1024 tracks.
-- 
Dave Remien - WINCO Computer Engineering Group (only somewhat confused, now)
208-526-3523 Paths: ...!bigtex!pmafire!dave | ...!ucdavis!egg-id!pma386!dave

mike@cimcor.mn.org (Michael Grenier) (12/16/88)

From article <450@lgnp1.LS.COM>, by phile@lgnp1.LS.COM (Phil Eschallier):
> to the best of my knowledge, AT style machine do not support disk with
> more than 1024 tracks.  i have read, however, that you can purchase a
> special ROM BIOS, and once you do this, you should be able to type the
> drive parameters in via fdisk and all should be well!

No. Unix doesn't use the BIOS as it isn't reenterent (and probably
doesn't run in protected mode anyway) except perhaps to load the boot
sector. My disk controller handles well over 1024 tracks... I have
no trouble formatting the drive to 1226 tracks.

I got ahold of someone at Microport today who thought they could look
into it in the next couple weeks and fix the thing. I am very happy
to wait a few weeks more...the RLL support was worth waiting for.

    -Mike Grenier
    mike@cimcor.mn.org

james@bigtex.cactus.org (James Van Artsdalen) (12/16/88)

In <659@bbking.KSP.Unisys.COM>, rmarks@KSP.Unisys.COM (Richard Marks) wrote:

> Hmmm . . . the basic PC disk IO interrupt 13 only has 10 bits for the
> track number.

But the underlying WD1010 register spec allows for 16 bits.
-- 
James R. Van Artsdalen          james@bigtex.cactus.org   "Live Free or Die"
Dell Computer Co    9505 Arboretum Blvd Austin TX 78759         512-338-8789

ken@uport.UUCP (Ken Chapin) (12/17/88)

In article <450@lgnp1.LS.COM> phile@lgnp1.LS.COM (Phil Eschallier) writes:
 >In article <618@cimcor.mn.org>, mike@cimcor.mn.org (Michael Grenier) writes:
 >> Does anyone know if Microport 386 or other versions of AT&T UNIX
 >> for the 386 handle disk drives larger than 1024 tracks? I currently
 >> have Microport 286 version 2.4 and the fdisk command doesn't
 >> support my drive which has 1226 tracks.
 >
 >
 >to the best of my knowledge, AT style machine do not support disk with
 >more than 1024 tracks.  i have read, however, that you can purchase a
 >special ROM BIOS, and once you do this, you should be able to type the
 >drive parameters in via fdisk and all should be well!
 
Our 386 3.0e product does support > 1024 cylinders and the 286 2.4 product was 
supposed to but we found a bug in our fdisk that wouldn't let us.  In any case,
the 3.0e 386 does and the 2.4 286 will very shortly support > 1024 cylinders 
without any special BIOS.

Ken Chapin         UUCP: ...!{sun | ucbvax | ihnp4}!amdcad!uport!ken
Microport Systems
Technical Support         

bernd@pfm.UUCP (Bernd Hennig) (12/18/88)

In article <618@cimcor.mn.org> mike@cimcor.mn.org (Michael Grenier) writes:
>Does anyone know if Microport 386 or other versions of AT&T UNIX
>for the 386 handle disk drives larger than 1024 tracks? I currently
>have Microport 286 version 2.4 and the fdisk command doesn't
>support my drive which has 1226 tracks.

I have the same problem with the Microport Unix 386 and a maxtor harddisk
with 15 heads and 1226 tracks. Only 140 megabyte of the 160 are available.



-- 
Bernd Hennig             uucp: uunet!mcvax!unido!sns!pfm!bernd
Oberer Moenchelenweg 2   PFM: *49 711 3700978 300/1200/2400 Bd 24h
D-7300 Esslingen              *49 711 3701830 300/1200/2400 08 am - 08 pm
.......a child of 5 could understand this......FETCH ME A CHILD OF 5 !!...

neighorn@qiclab.UUCP (Steve Neighorn) (12/18/88)

In article <618@cimcor.mn.org> mike@cimcor.mn.org (Michael Grenier) writes:
>Does anyone know if Microport 386 or other versions of AT&T UNIX
>for the 386 handle disk drives larger than 1024 tracks? I currently
>have Microport 286 version 2.4 and the fdisk command doesn't
>support my drive which has 1226 tracks.

AT&T Unix can have problems with disks that contain more than 1024
cylinders. This will be fixed in a Release 3.2 upgrade. For the time
being, you must set up the system with what is called "logical sector
translation." Here is an example of how it works using a Maxtor 4380
ESDI drive (1222 cylinders, 15 heads, 36 sectors/track), an Adaptec 
2322B controller, and AT&T Unix Release 3.2.

1) Install hard disk and controller in the machine. Set the Maxtor up
   for soft-sector operation.
2) Use setup to program the CMOS Drive Type for Type #1.
3) Boot a DOS disk.
4) load debug, and enter the command: g=c800:5
5) Choose selection 0 (format the drive) from the main menu.
6) Follow the instructions regarding bad sector mapping, and set the
   interleave for 1:1. When the format is done it will return to the
   main menu.
7) Choose selection 1 (DOS 3.3 style partition) from the main menu.
8) Partition the entire disk for partition #1. Return to the main menu.
9) Exit the BIOS formatting program.
10) Turn the machine off, take the Adaptec controller out and set the
    jumper that disables the controller's BIOS.
11) Use setup to program the CMOS Drive Type for a drive with the following
    parameters:

    Cylinders = 654
    Heads     =  16
    Sec/track =  63

    The trick here is getting the cylinder count under 1024 by increasing
    the other settings. You do lose a few hundred sectors with this technique.
    
    OLD: 1222x15x36 = 659880
    NEW:  654x16x63 = 659232

12) Boot your Unix disk and format as usual.

NOTES: Step 11 is the hard one for some machines, as not all 80386 machines
       have a BIOS that supports user definable drive types. 3.2 Unix will
       let you change the number of cylinders for a drive from within the
       installation program, but you must run "setup" to change the heads
       and sectors/track.

       The WD1007 gets around this limitation by providing user definable
       drive support in its own BIOS. (This is my understanding of the
       card - I don't have one myself).

       Unix 3.2 will appear to function, at least for awhile, if you
       ignore the above steps and just use the drive with 1222
       cylinders. But sometime during the installation of the system,
       perhaps the kernel rebuild with the 2K file system, the console
       will suddenly fill with BAD BLOCK remap messages. Then the
       dreaded BAD BLOCK IN UNIX SACRED AREA will appear. At that point
       you are in the unenviable position of being UPS (up shit creek).
       Yes, I tried this first, just to be sure :-(.

Good luck with your big disk!

-- 
Steven C. Neighorn            !tektronix!{psu-cs,reed,ogccse}!qiclab!neighorn
Intel Corporation            "Where we BUILD the Star Fighters that defend the
Development Tools Operation      frontier against Xur and the Ko-dan Armada"
80960 Language Group            work: (503) 696-7264 / home: (503) 645-7015

plocher@uport.UUCP (John Plocher) (12/18/88)

>> Hmmm ... the basic PC disk IO interrupt 13 only has 10 bits for track number.
>
>But the underlying WD1010 register spec allows for 16 bits.

But the WD-1002 controller only uses 10 bits, the WD-1003 uses 12.  So much for
specs... :-)

   -John Plocher

dave@micropen (David F. Carlson) (12/21/88)

In article <280@uport.UUCP>, plocher@uport.UUCP (John Plocher) writes:
> >> Hmmm ... the basic PC disk IO interrupt 13 only has 10 bits for track number.
> >
> >But the underlying WD1010 register spec allows for 16 bits.
> 
> But the WD-1002 controller only uses 10 bits, the WD-1003 uses 12.  
>    -John Plocher

Of course, the WD1002 is a floppy only controller.  For the time being
addressing 1024 cylinders (8.5 meg) on a floppy seems sufficient!  :-)


-- 
David F. Carlson, Micropen, Inc.
micropen!dave@ee.rochester.edu

"The faster I go, the behinder I get." --Lewis Carroll

plocher@uport.UUCP (John Plocher) (12/22/88)

+-- In article <595@micropen> dave@micropen (David F. Carlson) writes:
|  +-- In article <280@uport.UUCP>, plocher@uport.UUCP (John Plocher) writes:
|  |  +--  someone else said:
|  |  | > the basic PC disk IO interrupt 13 only has 10 bits for track number.
|  |  | But the underlying WD1010 register spec allows for 16 bits.
|  |  +--
|  |  But the WD-1002 controller only uses 10 bits, the WD-1003 uses 12.  
|  +--
| Of course, the WD1002 is a floppy only controller.  For the time being
| addressing 1024 cylinders (8.5 meg) on a floppy seems sufficient!  :-)
+--

Sorry, the WD 1002 is a combined hard disk / floppy disk controller that IBM
and most clones sold until the WD 1003 came out.

The 1002 has lots of discrete logic on board to implement both floppy and hard
disk support for the AT.  The 1003 is the same thing done with custom chip sets.

I know because I have several of each here in front of me, and my home system
(hobbes) has a 1002 in it.  I can't say how WD did it, but that "Floppy Only"
controller sure handles that 80Mb CDC Wren fine!  :-)

I know I sometimes stick my foot in my mouth (ok, both of them...), but this
once I know what I'm talking about. :-)))

   -John Plocher

joel@intelisc.UUCP (Joel Clark) (12/22/88)

In article <618@cimcor.mn.org> mike@cimcor.mn.org (Michael Grenier) writes:
>Does anyone know if Microport 386 or other versions of AT&T UNIX
>for the 386 handle disk drives larger than 1024 tracks? I currently
>have Microport 286 version 2.4 and the fdisk command doesn't
>support my drive which has 1226 tracks.

Procedure to access > 1024 tracks.  The 63 sector translation method.
on AT&T UNIX Sys V R3.1 or R3.2 as released by iNTEL  (or Bell Tech.)

I have never done this, but it has been done twice here by some other
engineers.  They never read this newsgroup so I will try to explain what they
have done.  This is also cryptically explained on the last 3 pages of the
"Adaptec ACB-2320/ACB-2322 User's Manual".  See also page 31 (General
Translation Mode) of the "WD1007A-WAH/WA2 Winchester/Floppy Disk
Controller" manual Document #79-000261 From Wester Digital.

You need a controller with low-level formatting BIOS.  We have used the
Adaptec 2322-B16 and the Western Digital 1007WA2-001

You need to use the formula:  
  Logical cylinders = Phys heads * Phys Sect PT. * Phys Cyl / 63 * 16

	Western Digital's formula is slightly different and explained
	in detail on pages 31 and 32 of their manual (see above reference).

You need motherboard BIOS that has or can create a drive type entry for
a disk with Logical number of cylinders, 16 heads, and 63 SPT

For example if you have a disk with 1224 cylinders, 15 heads, and 36 Sectors
Per Track, and you format this with the Adaptec low-level format you will
actually have 1222 cylinders, 15 heads, and 36 SPT available.  (2 are saved
for bad sectors)   Plug these into the formula you get:
654 = (15 * 36 * 1222)/ (63 * 16)
So you need a motherboard BIOS drive type entry for a disk with 654 cylinders,
16 heads, and 63 SPT.

Now execute the following steps:

1)  Boot DOS and run the controller format routine:
	A>  DEBUG
	-   g=C800:5

	For Adaptec pick '0' to primary format the drive.  format will inform
	you it is formatting a 1222 cylinder disk.  Continue.

	For WD we were using Drive type 1 in the motherboard BIOS.
	This specified 618 cylinders 16 heads and 63 SPT.  So we
	hit '+' until a disk type matching our logical cylinders etc.
	was shown.  Continue with format.

	Enter Defect information.

2)  For the Adaptec controller when formatting has completed, shutdown the
machine and disable the Adaptec BIOS by removing the BIOS enable jumper from
the controller.

3)  Reset the processor and run setup to pick or create the appropriate 
drive type (based on the logical cylinders, heads, and SPT)

4)  Insert the UNIX boot floppy and reset the system.  Install the UNIX
software but DO NOT REFORMAT THE DISK.



WARNING: CHANCES ARE AT BEST 50/50 THAT YOU WILL THE RIGHT MOTHERBOARD,
	 AND BIOS ON THE MOTHERBOARD AND CONTROLLER CARD TO ALLOW THIS
	 TO SUCCEED.

To do this on a WD controller on R3.0 UNIX we had to mod our motherboard.
To do this on an Adaptec Controller 2322-B16 we had to get new controller
firmware.

Best of Luck

Joel Clark
iNTEL Scientific Computers			joel@intelisc.intel.com
Beaverton, OR	(503) 629-7732		{tektronix}!ogccsu!intelisc!joel

howardl@wb3ffv.UUCP ( WB3FFV) (12/25/88)

In article <595@micropen>, dave@micropen (David F. Carlson) writes:
> 
> Of course, the WD1002 is a floppy only controller.  For the time being
> addressing 1024 cylinders (8.5 meg) on a floppy seems sufficient!  :-)
> -- 
> David F. Carlson, Micropen, Inc.

   Well this is not 100 percent correct, the WD-1002FOX is a floppy controller,
but the WD-1002WA2 is a combination HARD/FLOPPY controller. Actually the 
original IBM-AT machines all came with the WD-1002WA2 series controller, as 
I replaced a good many of them in my day. Just figured I would set the record
straight...


-------------------------------------------------------------------------------
Internet  : howardl@wb3ffv.ampr.org	|	Howard D. Leadmon
UUCP      : wb3ffv!howardl		|	Fast Computer Service, Inc.
TELEX     : 152252474     		|	P. O. Box 171 
Telephone : (301)-335-2206		|	Chase, MD  21027-0171