[net.unix] TS-11 Configuration Problem on 4.2BSD

davem@sdcrdcf.UUCP (David Melman) (07/21/86)

Keywords:

We recently purchased two Cipher-990 6250 bpi tape drives
and one Dilog DU142 tape controller, which can control up to
four drives. The DU142 emulates the TS-11.
This is installed on a Vax-780 running 4.2BSD.

The problem:  How to configure UNIX for two TS-11 drives.

The TS-11 driver man pages says: 
"only one transport is possible per controller".
However, the DU142 has separate csr addresses for each tape drive.
The first one is at csr 0172520, the second is at 0172526, ...
which makes me believe the kernel can be fooled into perceiving
the one controller as two.

I have the kernel configured with one drive as the man page dictates:
"controller zs0 at uba? csr 0172520 vector tsintr"
"tape ts0 at zs0 drive 0"
and the first tape drive works.

How do I specify the second drive?

I tried :
"controller zs1 at uba? csr 0172526 vector tsintr"
"tape ts1 at zs1 drive 0"
and got a crash when accessing either device.
-----------------
Any assistance will be most appreciated!

David Melman
System Development Corporation, Santa Monica
(213) 820-4111 X5367

UUCP: {sdcsvax,ihnp4}!sdcrdcf!davem

dan@rna.UUCP (Dan Ts'o) (07/23/86)

In article <2893@sdcrdcf.UUCP> davem@sdcrdcf.UUCP (David Melman) writes:
>However, the DU142 has separate csr addresses for each tape drive.
>The first one is at csr 0172520, the second is at 0172526, ...
>which makes me believe the kernel can be fooled into perceiving
>the one controller as two.
>
>I have the kernel configured with one drive as the man page dictates:
>"controller zs0 at uba? csr 0172520 vector tsintr"
>"tape ts0 at zs0 drive 0"
>and the first tape drive works.
>
>How do I specify the second drive?
>
>I tried :
>"controller zs1 at uba? csr 0172526 vector tsintr"
>"tape ts1 at zs1 drive 0"
>and got a crash when accessing either device.

	Are you sure that 0172526 is correct ? I have two Kennedy drives on a
Dilog DU132 and the addresses are 0172520 and 0172524. I believe they should
be 0 mod 4.
	Also, of course, make sure the /dev entries and the cabling is correct.
With the Kennedy's, the cabling is run in parallel!, like a bus.

					Cheers,
					Dan Ts'o
					Dept. Neurobiology
					Rockefeller Univ.
					1230 York Ave.
					NY, NY 10021
					212-570-7671
					...cmcl2!rna!dan
					rna!dan@cmcl2.arpa