"The Wizard of Oz" wrote in message news:pan.2004.08.30.18.42.43.260645@emeraldcity.gov... > On Thu, 26 Aug 2004 16:46:07 -0600, Laine Houghton wrote: > > > > > "The Wizard of Oz" wrote in message > > news:pan.2004.08.26.20.49.25.483509@emeraldcity.gov... > >> Quick question... First off I managed to do this 10 years ago and can't > >> find the program source. It must be on a dead hard drive. > >> > >> I want to be able to access the RAMRD and RAMWT switches from assembler. > >> I'm using 8bit Merlin (thanks for the manual Bill). According to an older > >> //e manual I have, the only thing I have to do is make sure 80STORE is > >> off. Regardless of the settings of 80STORE when I do a "STA RAMRD" > >> followed by a "LDA (somezeropage),Y" it looks as though there is a branch > >> occurs into page 1. Specifically into an area I have for tables. Has > >> anybody encountered this? Also can someone provide me with a quick code > >> segment they use to get these switches working. (FYI, I'm writing the > >> program on a ROM 03 GS and testing it on a ROM 01 GS with the eventual > >> users being //e and //c folks.) > >> > > > If your readaux is somewhat more complicated, assemble it for auxmem and > > read it into memory and use the $C311 auxmove to locate it in auxmem. in > > zero page have sta readaux jmp mycode. > > I took a look at your source and compared it to my own. I had the > same set up. In fact I had one of the equates defined twice. I just added > a line to make sure I was using the zero page in main memory and cleaned > up the initialization code. I've compiled it but haven't tested it. I've > been doing a bit of thinking. The code works elsewhere in the program. It > SHOULD work here. > > Maybe I'm not looking at a coding problem... I've got a bunch of > subroutines followed by some data. The whole thing is ORG'd to $4000 > because it won't fit in at $C00 - $1FFF with the tables. The test machine > is booting into GSOS then going into P8. Could I be running into some sort > of memory management issue? Could Applesoft be trying to use the area? > Suppose I ORG'd the code to $C00 and the tables to $4000... > > Lots of things to try. It's too late to get it ready for the 1'st. > Perhaps October 1'st. > > Later > Mike I understand you have your code and tables ORGed at $4000. I assume you bload or brun the program. this puts the code and tables in main memory. Are you also copying the code and tables to the same area of Aux memory? AUXMOVE EQU $ C311 ;carry=1 main>aux carry=0 aux>main STARTLO EQU $ 3C STARTHI EQU $ 3D ENDLO EQU $ 3E ENDHI EQU $ 3F DESTINATIONLO EQU $ 42 DESTINATIONHI EQU $ 43 LDA #0 STA STARTLO STA ENDLO STA DESTINATIONLO LDA #$40 STA STARTHI STA DESTINATIONHI LDA #$40+FILE_LENGTH STA ENDHI SEC ;MAIN TO AUX JSR AUXMOVE EXAMPLE: $4120 8D 03 C0 ;STA C003 $4123 B1 FE ;LDA ($FE),Y As soon as the STA C003 is triggered, the B1 FE at $4123 isn't seen. What ever resides at $4123 in aux memory is what gets executed. Also if FE is pointing to say $4200 the value loaded in the accumulator will also be from aux memory. If you have use of a windows machine and can accept a 900K file I can send you a debugger that will allow you to find the problem in about 4 minutes. Well, give it 15 minutes with a first time learning curve. Also you can develope the program in the emulator in a pure //e enviroment. Imagine compile time at 50 megahertz :) Or you can send a binary and I'll tell you whats tripping it up. Laine... Rich if you're listening, I tried to send you version 2.0 and it got bounced back to me. I used aiiadict@hotmail.com