[comp.sys.transputer] MAC II, LSC, and Transputers from Levco

huangch@eecs.nwu.edu (Chih-Hao Huang) (04/17/89)

Does any one out there have any experience with Light Speed C and Transputers?

We have transputer board (with T800 X 4) supplied by Levco on Mac II.
Long before we decided to use transputers, we developed a (large) system which
heavily depended on ToolBox and was written in LSC + Programmer Extender.
(This is the practical reason why we can not switch to MPW C now.)
Now we want to use transputers as part of the system. ( don't ask me why, I'm
not the boss.) We can use the development system (TCX, TLINK, TCNV ... etc,
under MPW environment ) to generate bootable codes (.B4 files) and run through
CIO. However, what we really need is to load the transputers from a host
program running on MAC II ( host program is executed by 68020 ). We intend to
write the host program in LSC. But we have not succeeded yet.

Here is what we have tried :

Take the program in "rice" folder as our subject to try.
"rice" folder contains a set of sample programs ( host.c, translink.h,
randomer.c, randomer.B4 ) which demonstrate how to load transputer from host,
and the host.c is written in MPW C.

(1) Correct the names of included files (DeviceMgr.h, FileMgr.h SlotMgr.h) for
	LSC.

(2) Change the sizeof(int) to sizeof(long). (In MPW C, "int" is "long"; however
	it is "short" in LSC.)

(3) Change the related the data structures from "int" to "long". (To confirm
	the type declarations in the header files.)

(4) Compile and link are O.K.

(5) Run it under debugger :

	Configuration is O.K. (We can check it by "Read004").
	Sending first few bytes (of the .B4 file) is O.K.
	"OutByte" always gets timeout when sending the 455th byte of the file.
	(its value is 0xAE, and the one before it is 0x22)

Here is the weird thing :

	We tried other .B4 files which do not contain any adjacent 0x22 and
	0xAE. They can be successfully loaded on to transputers and running.
	We also tried some .B4 files which do contain this pair, and they
	always get time out at the time when sending 0xAE right after 0x22.

Do we forget something?
Do we do anything wrong?
Does any one have similar experience?

Any information will be greatly appreciated. ( please give me a mail )

( No flames please, we are new to transputers :-)



******************************************************
*						     *
*	Chih-Hao Huang	(huangch@eecs.nwu.edu)	     *
*						     *
*     EECS Department at Northwestern University     *
*						     *
******************************************************