Re: TAC-3 to VME Box Interface

[prism!jm59@gatech.edu (MILLS,JOHN M.)]

In article <9304061817.AA03651@srs21g.scf.loral.com> mkm%srs21g@SCF28.SCF.LORAL
.COM (Mike K. McNair) writes:
>To Anyone Who Can Help,
 >
 >We will be developing a system which has basically two boxes, a TAC-3 (which
 >is essentially an HP workstation) and a VME card cage, connected by an Ethern
et.
 >Any ideas on which compiler/run-time vendors support Inter-Process, Inter-Ada
 >Task and/or Inter-Processor communications for this configuration? By the way
,
 >we will be using Motorola 680x0-based processor boards in the VME box.

We are integrating three motorola 68030 boards with a number of standard and
custom peripherals in a single VME chassis.  This box also communicates by
serial i/o with a second VME chassis using Radstone 60820-based cards.

We use the SD-Scicon XD-Ada compiler run in a VAX/VMS environment.  We use the
XD-Ada message-passing mechanism between tasks on the same and different
processors.  We debug using ethernet and the resident monitors on the MVME
processors.  Very happy with all these.  I understand there is a UNIX-based
version of this compiler, but I don't know if it's available for the HP host.

We may be pushing the limits of the message-passing utility; we're implicitly
invoking several activities through 'timed receive' and 'pend on message'
dependencies, and have sometimes observed problems synchronizing startup 
on our three processors.  I think this is due to the way our tasks were
written, and not necessarily to the message-passing mechanism.

We only use ethernet for monitor debugging, so I can't speak to it's use
in interprocessor task communications.

Don't know if this is close enough to your environment to be any help.

Regards --jmm--

-- 
John M. Mills, SRE; Georgia Tech/GTRI/TSDL, Atlanta, GA 30332
uucp: ...!{decvax,hplabs,ncar,purdue,rutgers}!gatech!prism!jm59
Internet: john.m.mills@gtri.gatech.edu
EBENE Chocolat Noir 72% de Cacao - WEISS - 42000 St.Etienne - very fine