[ Sorry for the delay, trying very hard to catch up on mail... --thorpej ]

On Mon, 17 Feb 1997 14:49:48 +0000 
 Peter Maydell <pm215@cam.ac.uk> wrote:

 > Looking through the archives, I see that HPIB interfaces can have up
 > to 8 'slaves'. Can somebody explain how this relates to the HPIB/GPIB 
 > 31 devices on an interface limit? Is the 'slave' concept only significant
 > to the existing hpib driver?
 > And shouldn't I have an hpib(4) man page?

Okay... 

The current HP-IB code really was written with CS/80 devices in mind..
I.e. the HP disk and tape drives that used an HP command set built on
top of IEEE 488.

Now, I'm not really much of a 488 expert (it's been a long time since I've
even _seen_ a 488-savvy Tektronix oscilliscope), so I probably won't
be able to answer your 488 questions...

Now, anyone wanting to implement HP-IB command passthrough may want to
take a look at arch/hp300/dev/ppi.c ... this is a "driver" for HP-IB
plotters... This is a legacy driver (i.e. inherited from Utah's HP-BSD),
and I don't have any documentation on the hardware... but I _can_ tell
you that the driver basically sends raw data from the user to the target
device... In this particular case, it would be HP plotter commands...
but it would be a good example of how to write a passthrough driver.

I don't really have a lot of time to actively work on a passthrough
driver, but if anyone wants to have their code reviewed, I'm happy
to do so, and when it's ready, include it in the standard NetBSD/hp300
sources.

ciao.

Jason R. Thorpe                                       thorpej@nas.nasa.gov
NASA Ames Research Center                               Home: 408.866.1912
NAS: M/S 258-6                                          Work: 415.604.0935
Moffett Field, CA 94035                                Pager: 415.428.6939