Subject: Re: status of MOP protocol?
To: Ken Wellsch <kcwellsc@math.uwaterloo.ca>
From: Thomas Riemer <triemer@babbitt.bernstein.com>
List: port-vax
Date: 05/23/1996 19:14:06
It wouldn't surprize me all that much if the VaxStations support bootp
in their boot ROM.  No, I'm not kidding.

I discovered that DecStations support bootp - at least the 2100 and the
5000 do.  If you read through the manuals, there is exactly one line that
indicates this - and only if you happen to know what you are looking for.

Seriously - search through a hardware manual about boot loading.

-Tom

On Thu, 23 May 1996, Ken Wellsch wrote:

> | > There's a mop daemon implementation by Terry Kennedy <terry@spcvxa.spc.edu> 
> | > at ftp://ftp.spc.edu/ but I don't know how good it is since I've never
> | > looked at it.
> | 
> | I just have looked at it, and it appears to be for RSTS/E, and not Unix.
> 
> What is nice is the V4 MOP manual in PostScript!
> 
> | Mats' one appears to do the job though. Now if I could convince NetBSD/vax to 
> | boot via MOP. Does anyone know what might be causing an SBC2INT error?
> 
> I experimented with Mats' mop software on a sun3 and got some nice info
> on the packet contents sent by my vs2000.  I did add a primitive part
> of BPF to the if_qe.c driver to allow me to do RARPD to boot a sun diskless
> from a vax.  But the MOP load request needs a registered multicast address
> or better to put the interface into promiscuous mode.  Rick gave me the
> lowdown on the very obscure way to put a DEQNA into that mode but I've
> not yet tried it, nor figured out how the PROMISC ioctl request makes it
> into the driver.  I also have no idea what the vs2000 loader does with
> an incoming image (in order to experiment with a "netboot" SA prog).
> 
> -- Ken
> 

---- Where theory and reality meet. 
---- Thomas Riemer