Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sdd.hp.com!ucsd!ucbvax!NASVM.BITNET!SWH From: SWH@NASVM.BITNET ("Steven W. Huggins") Newsgroups: comp.lang.asm370 Subject: Re: Long Programs ( > 4096 bytes) Message-ID: <9011052011.AA06457@ucbvax.Berkeley.EDU> Date: 5 Nov 90 18:44:06 GMT References: Sender: daemon@ucbvax.BERKELEY.EDU Reply-To: IBM 370 Assembly Programming Discussion List Distribution: inet Organization: The Internet Lines: 29 On Mon, 5 Nov 90 13:42:08 EST said: >Okay, simple question for those of you who do this sort of thing all the time: > >I am writing a program that has gotten a little long...long enough to lose >touch with the variables and literals. I have used the multiple base-registers >to offset this problem some, but it is still getting at me. One reason is >that I manipulate a matrix that is 4k long. Can anyone make the concept of >Control Sections and the method of dividing up a task so that each section >can reference the variables it needs in a common block make sense to me? > >AIIIGH! I hate being born in the age of Pascal and C....I happen to like >370 ASM.... Hi David, First thing to do is to move the matrix to the end of the program, after the literal pool, if you don't have a "LTORG ," statement add one before the matrix. Next change the way you address the matrix, use "L Rxx,=A(MATRIX)" instead of "LA Rxx,MATRIX", the LA might work for a while if the first byte of matrix is addressable. Control sections isn't really needed here. If you want more info on control sections give me a call, I hate typing. VSE, the stories of our demise are greatly exaggerated, VSE/ESA, Steven W. Huggins B I Moyle Associates, Inc. 703-803-8686 Office 612-933-2885 Home Office 202-334-3592 NAS Client(Mon & Wed)