[pups] Hardware config for 2.11BSD Installation on PDP-11/73

[Steven M. Schultz]

Hi -

> From: "Chuck Dickman" <chd at nktelco.net>
> The memory board that I have was made by Clearpoint Inc. The board

	'fraid I can't be of help with the memory board - that's not
	my area of expertise.

> I have two MSCP controllers installed. One is a CMD CQD-220/M SCSI

	Ah, but I can help with getting the dual MSCP controller
	question.

	My 11/73 is similar - it has an Emulex UC08 and the original
	RQDX3 (with an RD54 and a RX03).

> DU0 is ra0 and DU4 is ra12 in 2.11 BSD. If I change the jumpers on
> RQDX3 to unit 2, I boot at DU8, the first RX50, but I can no longer
> access the floppies from 2.11 BSD. Looking at the sources, it seems
> that the drive number is limited to 0-7, but the controller seems to
> want unit number 8. What am I missing? This seems to occur both in
> the kernel and in the standalone installation system.

	Ignore anything you might know about DU numbers ;)

	2.11 numbers the drives 0 thru 7 on _each_ controller.   Thus
	'ra0' thru 'ra7' are on the first controller, 'ra8' thru 'ra15'
	on the second controller and so on.  There is a maximum of 4
	controllers supported (two bits in the minor device number).

	The standalone I/O system uses device names of the form:

		dev(ctlr, unit, part)

	where 'ctlr' is the controller number (0 thru 3), the 'unit' number
	is 0 thru 7 and the 'part' is the partition number (0 thru 7).   So,
	to access the first drive on the second controller from a standalone
	program the name would be:

		ra(1,0,0)

	you probably will be prompted for the CSR since 'boot' and friends
	only know about the first controller's address.

	Once the kernel is loaded you access drives on the 2nd controller
	with device names such as /dev/ra8a, /dev/ra9a, and so on.   

	The "sparse" numbering was chosen (eons and eons ago) to make the
	MSCP disks look more like the traditional disc controllers which
	allowed for 8 drives per controller.

	In my /etc/dtab file I have:

ra      ? 172154 154    5       raintr          # uda50, rqdx1/2/3
ra      ? 172150 0      5       raintr          # 2nd uda50/uc08

	From the console ODT prompt (I've an older MXV11 boot set up) I
	use "172150 du 0" to boot from the SCSI disk attached to the UC08.
	When the system is up the RD54 (which is unit 0 on the other MSCP
	controller) is accessed as 'ra8' (and the RX03 which is unit 1 
	accessed as 'ra9').

	It's pretty simple and regular once one knows what is going on ;)

	Steven Schultz
	sms at to.gd-es.com

Received: (from major at localhost)
	by minnie.cs.adfa.edu.au (8.9.3/8.9.3) id MAA62645
	for pups-liszt; Mon, 5 Feb 2001 12:52:20 +1100 (EST)
	(envelope-from owner-pups at minnie.cs.adfa.edu.au)