Post by thread:No-assemble code

Hi again, Here's a real live example of Microchip code that won't assemble in the Parallax PASM and I don't know why. The error I get is in the first line of the "test program". The error says: "data was already entered at 000h" and the line that generated it is: "main movlw 0F3h" So ... what data was already entered? There is no mention of a location 000h in the first place. Perhaps this is one of those instances where the assembler is totally confused by something else, but prints that as the error. You know how these software things are. Any ideas? Tnx, Jim Johnson spam_OUTjjohnsonTakeThisOuThpl.hp.com ;******************************************************************* ; include "mpreg.h" include 'dbl_add.asm' ; JJ - attempt at including a file include 'dbl_jj.asm' ; JJ - another include file ; org 0 ; LupCnt equ 10 ; Number of iterations ; SqrtLo equ 10 ;ACCaLO SqrtHi equ 11 ;ACCaHI ; NumLo equ 1Dh NumHi equ 1Eh count equ 1Fh ; ; init movlw LupCnt movwf count movf NumHi,W movwf SqrtHi movf NumLo,W ; set initial guess root = NUM/2 movwf SqrtLo bcf STATUS,CARRY rrf SqrtHi rrf SqrtLo retlw 0 ; div2 bcf STATUS,CARRY rrf 13,W ;ACCbHI movwf SqrtHi rrf 12,W ;ACCbLO movwf SqrtLo retlw 0 ; Sqrt call init sloop movf NumLo,W movwf 12 ;ACCbLO movf NumHi,W movwf 13 ;ACCbHI ; call D_divS ; double precision division call D_add ; double precision addition ; ; the above 2 routines are listed ; ; as seperate routines call div2 decfsz count goto sloop goto over ; all iterations done ; ; branch back to desired location ; ;************************************************************* ; Test Program ;************************************************************* ; main movlw 0F3h movwf NumHi movlw 0F6h ; Set input test number = 62454 movwf NumLo ; = F3F6h ; goto Sqrt ; cannot use CALL : Math routines ; ; use up all the stack. over nop ; all iterations done ; self goto self ; result = 00F9h = 249 ; ; exact sqrt(62454) = 249.9 ; org 0 goto main ; END

Jim, your specific problem is this: At the beginning of your code it starts out: >> init >> movlw LupCnt >> movwf count >> ... Now you've commented out the org 0 so that was ignored, however in the absence of any directive the assembler starts loading code at location 0. (makes sense yes?) Now later in your program you have: >>; >> org 0 >> goto main >>; >> END Which re-sets the program counter to zero and trys to assemble the 'goto main' into that address. Which overwrites the 'movlw LupCnt' that you had in init above. The correct fix for this specific problem is that you should move the 'goto main' to be the first statement in your program and then it will assemble at location 0 and all will be well. I both agree and disagree with what has been said about Parallax's assembler. On the one hand, if you've got working MPASM code it can be a real pain to get it to assemble correctly with the Parallax assembler. On the other hand, if you learn the Parallax assemblers quirks early on (and MPASM has its own) then writing new code from scratch can be fairly easy in both. The lack of a macro facility and conditional assembly in the Parallax assembler hurt it, however if you are using Linux then you could always run cpp over the code before feeding it to pasm. --Chuck McManis All opinions in this message/article are Sun Microsystems Inc. those of the author, who may or may not Internet: .....cmcmanisKILLspam@spam@Eng.sun.COM be who you think it is. Crypto-puzzle: *0U0JPFPrWRN9PkWRKeP5WRmIR9wP5QAWuIQP9Pu9tnIZ7AD1SIS