Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Path: utzoo!utcs!mnetor!seismo!mcvax!ukc!cstvax!hwcs!hwee!hmc From: hmc@hwee.UUCP Newsgroups: net.lang.pascal,net.wanted.sources,net.unix,net.unix-wizards Subject: Berkeley Pascal for Sys V Message-ID: <126@hwee.UUCP> Date: Mon, 14-Jul-86 06:59:46 EDT Article-I.D.: hwee.126 Posted: Mon Jul 14 06:59:46 1986 Date-Received: Wed, 16-Jul-86 08:45:21 EDT Reply-To: hmc@hwee.UUCP (Hugh Conner) Organization: Elec. Eng., Heriot-Watt U., Edinburgh, Scotland Lines: 32 Xref: utcs net.lang.pascal:571 net.wanted.sources:2556 net.unix:8603 net.unix-wizards:18181 We are running a Perkin-Elmer 3230 under PE's Xelso operating system, which is standard System VR2. I have been trying to put up the Berkeley Pascal on this system. We have Release 3.0 of this which is written for 4.2BSD on a Vax or 68000. I decided that the interpreter would be easier to port than the compiler. I have come across two major problems. 1) The interpreter uses the machine's stack to pass arguments to routines. There is a sed script for modifying the assembly code of the interp routine to do this. 2) It assumes that longs can start on a short boundary. I solved the first problem by just using a global array for the stack and not trying to use the machine's stack. I have been gradually sorting out all the boundary problems but it is a slow and tedious process. Has anyone gone through this loop before and produced a genuinely portable version of Berkeley Pascal? I would be interested in either/both the interpreter (pi/px) and the compiler (pc). Sorry about posting this to so many groups, but I hope that way I might get lucky. -- *+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++* + + + "Who are all these people in my office anyway?" + + + + Hugh M. Conner hmc@ee.hw.ac.uk + + + *+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++*