🚀 go-pugleaf

  I have collected all of the mail and postings I assembled from
people either interested in or doing work on various forms of free
transputer software.
  I would like to hear more about all of these topics, so if this
should spark more discussion, I would appreciate being cc:ed.
[[[ Please respond to bray@wcuvax1.wcu.edu, NOT dbt@adm... Thanks.]]]

Happy Hacking,

--Jim

From: Alessandro.Forin@ERNST.MACH.CS.CMU.EDU
Subject: RE: Mach on transputers?
Content-transfer-encoding: 7BIT

I think it is a great idea.  As far back as two years ago I have
encouraged the Grenoble folks to look into this, esp given the
support that chip was getting in Europe.  I do not know if they
have followed my advice or not.

It does seem that the still-not-ready T900 (?) will not win that big
over say alpha-based machines.  But as you point out there are other
interesting properties in that chip.
The lack of an MMU is, on the other hand, pure brain damage.

sandro-

From: michael@robots.oxford.ac.uk (Michael Stevens)
Subject: Mach on Transputers?
Content-transfer-encoding: 7BIT

Jim,
	If you do get any info about people doing Mach3 on a
transputer, I would also be interested. I would be quite happy to help
out if someone is working on such a project.

Michael Stevens, Robotics Research Group,Dept of Engineering Science,Oxford,UK
INTERNET: michael@robots.oxford.ac.uk

From: dlb@osf.org
Subject: RE: Mach on transputers?
Cc: mach3@cs.cmu.edu
Content-transfer-encoding: 7BIT

>   Has anyone given thought to running Mach3 on transputers? While I

To some extent.  The big drawback is that the current parts (T800
series) have neither virtual memory nor protection of any sort.
The next series of parts (T9000) will have physical segmentation
with segment based protection.  The OSF Research Institute is working
on a modified prototype of Mach which should work on such an
architecture (segmentation, but no page based virtual to physical
memory mapping).  There are currently no plans to port this to
transputers, but if someone wants to volunteer ...

--Dave

Subject: [u31b3hs@cip-s01.informatik.rwth-aachen.de: Code for accessing a transputer link under Linux 0.96a available per ftp]

I got too much requests, so now you can get my first work from
ftp.thp.uni-koeln.de, path pub/linux/transputer, file sirius.tar.Z.

It contains a user-space link driver (hopefully) inmos specs compliant,
source code for a small link test program, a transputer binary and a
transputer disassembler.  If you have iserver, you should be able to
link it with my driver.

I don't make any warranties, releasing that stuff is to reduce my
e-mail.  Feel free to suggest improvements, bugfixes.  *This release
does not contain the assembler/linker.*

Future releases will follow :)

Michael

[[[ I never was able to reach his machine because of network timeouts. -jb ]]]

From: in%"mwe@isy.liu.se"
Date:  1-SEP-1992 13:47:22.63
To: BRAY
CC: 
Subject: RE: Good C++ or C Compilers? reply


I am currently working on a port of gcc/g++ to the transputer. The current status
is that I have a working as/ld and a gcc that generates something that does compile
and load, but does not work.
The time I can spend on the project is rather limited, so I dont know when
it will be finished. If you want to look at what I have I can make it available
to you at our ftp-site.

Mikael Wedlin                                               Phone: +46-13282707
Dept of Electrical Engineering                         Internet: mwe@isy.liu.se

[[[ I have not heard from Mikael in a long time. I hope I hear more
about this if it progresses. --jb ]]]


From: gkt@iitmax.iit.edu (George Thiruvathukal)
Subject: Transputers and Linux
Organization: Illinois Institute of Technology / Academic Computing Center
Date: Thu, 15 Oct 92 15:51:00 GMT

Is anybody out there using Linux to develop transputer software? If so, please
send me information about what you are doing and what tools you are using to
get it done.

For some time, I have been interested in porting gcc as a cross compiler for
the Inmos transputer family. Since I use T805-based boards from Computer
System Architects, my inclination is to cross compile to the T805. The T805
supports fixed and floating point and is a RISC processor. As these transputer
boards do not directly support I/O, an I/O server must be written to perform
I/O on the host processor, e.g. a PC running Linux.

If there is anybody out there interested in the possibility to work on a free
Linux-based toolset for transputers, or if someone is doing it already, please
let me know.

I will post a summary on this subject. Hopefully, there will be some responses
to warrant a summary.

-- 
George K. Thiruvathukal

IIT - School of Computer Science - gkt@iitmax.iit.edu
Tellabs - Data Communications Division - att!tellab5!gkt