[net.unix-wizards] Benson plotters and 4.1 BSD

dbb@helens@sri-unix (06/27/82)

Date: 10 Jun 1982 14:59:17-PDT
I designed the (new) DMA interface for Benson and participated in converting
several systems with Benson plotters at Stanford to 4.1 BSD.

The latest revisions for the Benson interfaces, so far as I know are:
	PIO 978600-00 board rev. S	schematic 978598 rev. L
	DMA 980336-00 board rev. D	schematic 980334 rev. D

There are three drivers in circulation:
	driver A:	This is used by Berkeley and distributed with 4.1BSD.
			It uses polling rather than interrupts and filled with
			code for watchdogging the hardware.  It won't work at
			all with the new DMA.  It apparently works at Berkeley
			with the OLD interfaces they have. 

	driver B:	This is used at several (at least 4) installations at
			Stanford.  It is derived from driver A, uses interrupts,
			and works fine with the new DMA.

	driver C:	This is used at Stanford Geophysics.  This driver was
			originally written by Rob Mathews and Peter Eichenberger
			for 3 BSD and was (partially) converted to 4.1 BSD.
			It works fine with the new DMA.  It is clean code,
			allows either DMA or non-DMA operation, but does not
			have all the features of 4.1 BSD drivers (autoconfigure,
			multiple devices, etc).

On our system (SU-HELENS) we are running an early version of the new DMA and 
a version of the PIO board (schematic rev. K) which has a known bug (fixed 
in schematic rev. L) that is not triggered on our system.  We run driver B to
stay as close to the stock 4.1 BSD as possible.  We have received the latest
rev. boards and will try them some day.

Stanford Geophysics has an AP-120B array processor which seems to find all of
the holes in the other hardware and software on the system.  When Geophysics 
converted to 4.1 BSD, they had a PIO board like ours (schematic rev. K) 
and found that their AP triggered the above-mentioned bug when running 
driver B, but not driver C.  Naturally, they have been running driver C ever 
since.

The only bug I know of in the latest rev. boards is that powering off the 
plotter during a plot will cause "nexus stray intr ipl14" messages on the
console.