The Atari Assembler Editor Reference By Matthew J. W. Ratcliff Released 20-NOV-89 I teethed on the Atari Assembler Editor (Asm/Ed) cartridge way back in 1982. Now, nearly seven years later, it is becoming more popular than ever. Why? Well, I have seen several mail order ads over the past year offering Asm/Ed for only $10 or even as little as $5. Here's the hitch, no documentation. A lot of Atari cartridges were left over from the Warner days, when the Tramiel family took over the company. It seems that thousands of cartridges were sold "by the pound", with no boxes or documentation, just to clear out the warehouses. Now many of you Atarians have decided that Asm/Ed was too good to pass up at such a low price, because you just know that someday you will learn assembly language programming. This article is for you, the Atarian who took the plunge and bought the Asm/Ed cartridge without documentation. I won't pretend to teach you how to write your own assembly language programs (although I'm bound to toss in a bit of free advice along the way), our Boot Camp series will do that job well enough. I hope to give you a quick comparison between BASIC and assembly language, to illustrate the major speed differences. I will also cover the mechanics of writing a USR routine to help speed up BASIC programs and a stand alone assembly program that may be executed from DOS. The Asm/Ed cartridge will work with any DOS (disk operating system) from Atari DOS 2.0s all the way through DOS-XE and even the SpartaDOS X cartridge. The Asm/Ed cartridge itself is made up of three basic components, the Editor, Assembler, and Debugger. I will present a quick reference for all the commands of each section of the cartridge, and then lead you through the creation of your first program. After booting your Atari with Asm/Ed installed, and control is sent to the cartridge, the EDIT prompt will appear. From here you can begin entering your assembly "source code", with line numbers, assembly mnemonics, and comments. The editor is line oriented, requiring line numbers. They are used for reference while editing only, and are not used as part of the program itself, as in BASIC. Editor NEW The NEW command clears all assembly source code from memory, providing a clean slate for entry of a new program. NEW will irrevocably erase your program from memory, so always be certain to save important code before using this command to start something else. I highly recommend using a comment as the first line of every program, similar to the following: 10 ;LIST#D:USRTEST1.ASM It is easy to forget what file you are working on, or make a typographical error when using the LIST command. Saving a file to the wrong place can ruin a lot of work in a big hurry. By placing the LIST command, followed by the correct file name, in a comment, you may simply displasy that line and use the full screen editor to eliminate the line number and comment character (the semicolon), press RETURN, and execute the proper save command every time. This good habit has saved me countless hours of frustration. Use it! DEL This command DELetes lines of code. The format is: DEL xx - Delete line xx from the program. DEL xx,yy - Delete lines xx through yy from the program. NUM NUMber lines automatically, for fast entry of source code. If no starting number is specified, the program begins with the last line number currently in your program, plus 10. Type your code and comments, pressing RETURN when each line is complete. Press RETURN only to stop auto-entry of code. Do NOT use full screen editing functions to change lines previously entered, while still in the NUM mode. This will give unpredictable results. The command format follows: NUM - Increments by the number 10 after each line. NUM nn - Begin line entry at line nn,increment by 10. NUM nn,mm - Forces the next statement number to be nn and the increment to be mm. The last command format may be used to insert new lines between code that already exists (e.g. NUM 11,1 to enter up to 9 lines of code between lines 10 and 20). REN RENumber the file. It will resequence all the line numbers of the source file. The command format is below: REN - Renumber starting with 10, increment by 10. REN nn - Renumber in icrement of nn, start with 10. REN nn,mm - Renumber by nn, starting with mm. You might think that you can load a LISTed BASIC program, and use this to renumber it. In fact you can, except that all the GOTO and GOSUB statements will not be resolved. For very small BASIC programs you can do this, but you are better off using a renumber utility. When inserting a lot of new code, REN may be used to space the line numbers wide apart, thus allowing entry of more new code with the NUM entry method between consecutive lines of source code. This is a bit archaic by today's standards, but works well. Should you move on to IBM PC based assemblers, or other sophisticated machines, you will find no such animal as a line number. You simply create and edit your source code in a word processor like environment. The assembler can easily resolve all the source code and labels without the need for those pesky line numbers. Asm/Ed and Mac/65 users must continue to deal with them, however. FIND The FIND command can be used to help locate any string of text anywhere in the program. This is the command format: FIND/THIS/ - Find the first occurrence of the characters "THIS", and display the line it is found on. FIND/THIS/A - Find all occurrences of "THIS". Each line is listed to the display as it is located. Pressing control-1 will throttle screen scrolling. FIND/THIS/xx - Find the string "THIS" on line number xx. The line is listed, if found. FIND/THIS/xx,yy,A - Find all occurrences of "THIS" from lines xx through yy, inclusive. Note that when searching, the line numbers themselves are ignored (you don't search the line numbers, when looking for a particular number). In the examples above the string of interest was delimited by the slash (/) character. Any matched pair of characters may be used as delimeters. The following would be used to find all occurrences of the slash character in your program: FIND */*,A REP REPlace strings in the file: REP/OLD/NEW/ - Replace the first occurrence of "OLD" with "NEW". REP/OLD/NEW/,A - Replace ALL occurrences of "OLD" with "NEW". Use the All option with care. REP/OLD/NEW/xx,yy - Replaces the first occurrence of "OLD" with "NEW", in the line number range xx to yy. REP/OLD/NEW/xx,yy,A - Replace all occurrences of "OLD" by "NEW" in lines xx through yy. REP/OLD/NEW/xx,yy,Q - Replace with query. You will be prompted to press Y for each replace. File Commands The following commands may all be associated with files: LIST - Display or save lines of source code PRINT - Same as list but omit line numbers ENTER - Retrieve source program SAVE - Save an object program (assembled code) LOAD - Load an object program In BASIC filenames are enclosed by quotes, such as SAVE"D:MYFILE.BAS". In Asm/Ed the filename is preceded by the pound sign (#); no quotes are used. Filenames you may use are shown here: #E: - The screen editor, used by default with some commands such as LIST. #P: - Refers to the printer. #C: - This is used in reference to the Atari program recorder. #Dn:filename.ext - This is a disk file. The 'n' refers to drive number, which may be from 1 to 8, depending on the DOS and drive configuration you employ. If no drive number is specified, drive number 1 is defaulted. The name of the file may be up to 8 alphanumeric characters, followed by a period and optional 3 character extender. The extender may be anything you wish. ASM or SRC is generally used for assembly source files, and OBJ or COM for executable object files. LIST The LIST command is used to display, print, or save to a file assembly source code. The formats are: LIST - List the entire program to the screen. LIST nn,mm - Display lines nn through mm. LIST #P: - Print the entire source program. LIST #C: - List the source code to cassette. LIST #D:FL.ASM - List the entire program to FL.ASM on disk drive number 1. LIST #D:FT.SRC,10,100 - List lines 10 through 100 to file FT.SRC. Any filename or device specification may be followed by the line number range specification. The PRINT command functions exactly like LIST, except that the line numbers are not output. Since assembly source files REQUIRE line numbers, it won't be very useful to PRINT your program to disk and attempt to ENTER it later. This would hopelessly confuse Asm/Ed; always LIST source code to disk. ENTER The ENTER command is used to retrieve a previously LISTed source file. A valid input device must be specified such as: ENTER#D:MYFILE.ASM ENTER#C: If you wish to merge a program, append a ",M" to the ENTER command, such as: ENTER #D:ROUTINES.ASM,M This merge works the same as the following sequence would in Atari BASIC: LOAD "D:MYPROG.BAS" ENTER "D:NEWSUBS.LST" The lines are merged. If any line numbers in the file to be merged match those of the file already in memory, the merge file takes precedence. If you wish to append a file to a current working program it may be best to ENTER the merge file first, RENumber it with some large range such as 20000,1 and then LIST it out to a temporary file. Then ENTER your main program, and finally enter with the merge option this renumbered file. SAVE Use the SAVE command to write a block of memory, such as an object program, to a file. Let's say your program begins at $4000 (with an *=$4000 at the top of your assembly code). After the ASM command, you see the final address was $41FE. Then, to create a binary image of this file, which may be loaded and run later, enter the following command: SAVE #D:MYFILE.OBJ < 4000,41FE Note that the addresses are always assumed to be in hexadecimal, and you do not specify a dollar sign ($) to indicate this on the SAVE command line. You may also SAVE to the cassette (#C:). With the proper ASM command, your object files may be created automatically, as we will see. Note that you may assemble your program in memory, and then go to DOS and use the DOS memory save command to create this object file as well. (The advantage of the DOS memory save command is that you can specify the RUN address as well, so that your program automatically executes when you perform a binary load. There are ways to set this up with the ASM command as well.) ASM Once you have created your assembly source code and LISTed it to a file for safe keeping, it is time to assemble it. This is Asm/Ed's primary function, to convert your source code into executable object code. When you issue the ASM command, the current file in memory is scanned for syntax errors. If it understands all your source code, all the assembly mnemonics are converted into equivalent binary codes and written to memory or a file. Care must be taken that your code assembles to an area of memory which does not conflict with your source code. Before assembling a program for the first time always enter the SIZE command. Three hexadecimal numbers will be displayed, such as: 10F4 1345 9C1F The first number indicates where in memory your source code begins, just above DOS's basic memory requirements. The second number is where, in memory, your source code ends. The final address is the top of useable RAM. At the top of your program will always be a statement similar to the following: 0 *=$4000 This tells the assembler to start building your object code at memory location 4000 hexadecimal, the program origin. This address may be any number between the second and third numbers reported by the SIZE command, with some notable limitations, when assembling to memory. It may be any value above the first number, so long as you assemble to a file. If you wish for your program to assemble into lower memory, you may use the LOMEM command. This must be the first command entered, once you start up Asm/Ed and receive the EDIT prompt. The format is: LOMEM xxxx, where xxxx is the hexadecimal address to set the new low memory value. This is the first address value reported by the SIZE command detailed above. For example, if you want your program to load at address $2400, and you know the object code will be 4K or less, then use LOMEM $3400 ($1000 is 4K bytes). Then ENTER your program, and use *=$2400 at the top of the file to set the origin. Then the program may be assembled in RAM to RAM safely, so long as your object code does not grow beyond 4K. If you plan to write stand alone assembly programs, which may be loaded from DOS with the binary 'L'oad command, I recommed an origin of $3400. This will set the start address of your code above both DOS.SYS and DUP.SYS RAM in Atari DOS, any version through DOS-XE, as well as any version of SpartaDOS. Unlike BASIC, you must manage memory yourself. If your program origin is too close to the second number from the size command the assembler may get confused. The assembler must build a symbol table and assign some temporary storage as it processes your source code. It starts building this information from the end of your source code, and grows upward. If the symbol table runs into the area where the object code is being stored in RAM, the assembler is likely to generate a lot of erroneous PHASE errors. If your origin is set too high, your object code will run into display memory and eventually out of room. These problems may be avoided in several ways. The general form of the ASM command is: ASM #D:SOURCE, #D:LIST, #D:OBJECT The first filename in the ASM command represents the file your assembly source code is stored in. This allows you to assemble from disk (but not cassette, since Asm/Ed requires multiple passes through the file). If this field is empty (simply place a comma immediately after the ASM command), then the source code is assumed to be in memory. The second filename specifies a listing file, where a complete "assembled listing" is routed. This will usually be the printer (#P:). If this field is left empty, the listing goes to the screen. The listing always goes somewhere; it cannot be turned off as it can in Mac/65. However, assembler "directives" may be used to control the output of a listing, as we shall see. The third field is the filename where the object code will be stored. If this field is not specified, your program is assembled to memory. Always make a current listing of your program on disk or cassette before issuing the ASM command. If you have set up memory mapping improperly, the source code will get clobbered in a big hurry. Assembler Directives Directives, or pseudo operations (pseudo-ops), are special instructions to the assembler. They can be used to control listing format, program title for listing, allocation of memory, and more. In general, assembler directives begin with a period (.), followed by some key word and associated parameters. OPT The OPTions directive controls assembler output. They are as follows: .OPT NOLIST - Suppress output of listing during assembly. .OPT LIST - Output assembly listing, default. .OPT NOOBJ - Do not generate any object code during assembly. .OPT OBJ - Output object code, default. .OPT NOERR - Do not display error messages while assembling. There is no good reason to ever use this option. .OPT ERR - Display error messages when assembling, default. .OPT NOEJECT - Do not skip 4 lines at the bottom of each page, when outputting the listing. .OPT EJECT - Skip 4 lines at the end of each page, default. More than one option may be placed on a single line, such as .OPT NOLIST,NOOBJ. Note that the Mac/65 assembler defaults to .OPT NOOBJ; it does not generate object code unless explicitly told to with the .OPT OBJ directive. Asm/Ed is just the opposite. Whenever you are assembling your program frequently, working out syntax and undefined label errors, it is generally wise to have a .OPT NOOBJ near the top of your program. When you are ready to generate code and start test running it, then change it back to .OPT OBJ. Title and Page The title and page directives are designed to make your assembly listings easier to read. The title directive is generally used to specify the name of your program, revision, and date. The page directive can be used to force a page break and optionally output some text. For example: 10 .TITLE "Attack Of The Killer Dweebies, Version 1.0" 20 .PAGE "Program equates" ... ... 200 .PAGE "Graphics Routines" ... ... 300 .PAGE "High Score Routine" Tab The TAB directive is used to set the spacing of the fields of your assembly code for listings. The command format is: 10 .TAB 12, 17, 27 The above example illustrates the defaults used by Asm/Ed. These may be set to any position you find most suitable for your printer listings. The first number indicates the field where the mnemonics (assembly opcodes) will appear, the second for the operands, and the third for the comment field. For example, suppose your program has a lot of long labels with a maximum of 15 characters. Then you may wish to set the tabs as follows: 10 .TAB 20, 25, 40 which would make for a prettier listing on the printer. BYTE, DBYTE and WORD The BYTE, DBYTE, and WORD directives are used to reserve storage in memory, similar to variables in BASIC. Labels may be associated with these directives for easy reference. For example: 100 LDA STORAGE 110 LDX STORAGE+1 ... 500 STORAGE .BYTE 34, $45 In the above the statement at line 100 will fetch the first value at location STORAGE, which is the number 34 following the BYTE directive. In line 110 the X register will receive the data value 45 hexadecimal. Note that the assembler will perform the address arithmetic "STORAGE+1" automatically. The BYTE directive may also be used to reserve storage for strings, such as: 100 LDA #STRING/$100 110 LDX #STRING&$FF ... 320 STRING .BYTE "This is a test", 155 In line 320 the BYTE directive reserves storage for a string initialized to "This is a test" followed by a 155 (ATASCII carriage return). The code in lines 100 and 110 fetches the address of the label STRING, placing its address high byte in A and low byte in X. This technique will be commonly used to pass the address of data collections (such as strings or data tables) to subroutines. The DBYTE directive reserves two consecutive memory locations, generally used for numbers greater than 256, in high byte, low byte order. For example: 1000 DATA .DBYTE 258 The above will result in two bytes of memory being reserved at location DATA, with the values 1 and 2 respectively (1*256 + 2 equals 258). Addresses are stored in low byte, high byte format as expected by the 6502 microprocessor. The WORD directive is used for this, such as: 100 *=$3400 110 START LDA #45 ... 290 RTS ;End of program 300 *=$2E0 310 .WORD START In line 100 the origin of the program, or program counter, is set to 3400 hexadecimal. The first line of code, with the label START, will then be assembled into your computer's memory at $3400. At line 300 the program counter is reset to $2E0. At line 310 we have the WORD directive, immediately followed by the label START. The assembler will "backtrack" as it processes your source code, realize that START refers to memory location $3400, and place this value (low byte, high byte sequence) in memory at $2E0, $2E1 respectively. This is a special location, $2E0, commonly referred to by name as RUNAD in Atari memory maps. When you assemble a program to disk, which will be loaded and run from DOS, you use the above technique to set the run address of your program. When the program ends with an RTS, control is returned to DOS. Most game programs do not end, but you will use this technique for many utilities. As we will see later, a BRK instruction is used, instead of RTS, when testing programs from Asm/Ed's debugger. Label Directive You do not have string, integer, and floating point variables in assembly language, the way you do with Atari BASIC. As we saw above, you must set up your own storage and interpret it properly. There are no automatic mechanisms in assembly language for managing variables. To make life easier, you will want to attach meaningful labels to constant values, such as: 10 RUNAD = $2E0 ... 1000 *=RUNAD 1010 .WORD START It is much easier to tell from this example that the intended run address of our program is defined at the label START. In the previous example for the WORD directive, we simply had the number $2E0. Unless you want to memorize a lot of memory locations, employ meaningful labels wherever practical. Labels are used for reference when you want to GOTO (JMP) or GOSUB (JSR) in assembly language. For example: 10 PROMPT .BYTE "PRESS RETURN TO CONTINUE",155 ... 100 LDA #PROMPT/$100 110 LDX #PROMPT&$FF 120 JSR PRINTSTRING ... 500 PRINTSTRING STA ICBADR+1 ;Print the string pointed to 510 STX ICBADR ;by A & X registers ... Labels may also be used in branch instructions such as: 50 CONTINUE LDA TABLE,X ... 100 DEX ; Decrement our loop counter, x register 110 BEQ EXITLOOP 120 BNE CONTINUE 130 EXITLOOP STA RESULT ... In the above example we have set up a loop, similar to a BASIC FOR/NEXT loop, between the labels CONTINUE and EXITLOOP. In line 100 the X register is decremented by 1 (we assume it was initialized by some code previous to line 50). If the result of the DEX instruction is zero (BEQ) then control will be passed to EXITLOOP. If the X register has not gone to zero (BNE) the control is sent back up to CONTINUE. As a result of DEX the zero flag can only be set (BEQ) or cleared (BNE), so we have exhausted the possibilities. It would have been equally valid to use: 120 JMP CONTINUE Generally, whenever you have the choice between a JMP and Bxx (branch) instruction, use the branch. It will require less memory and work faster. The problem is that a branch is limited to plus or minus about 127 bytes from the current position. If you try to branch too far, you will get an assembly error. Then JMP instructions, or combinations of JMP and branch instructions may be required. Origin Directive We have already used the "*=" directive in many of the previous examples. This tells the assembler to "set the program counter to the following address". The address may be some number, or a label, or some expression (so long as the assembler may resolve it to a fixed value). Some examples are: 100 *=$3400 ; Always a safe place to start a program ... 300 START = $4400 310 *=START ... 500 *=START + 439 ... 600 HERE *=*+45 ; Reserve 45 bytes of storage at HERE Take note of the spacing used in all of our examples. Any label always begins one space after the line number, referred to as the label field. The opcode field begins at least one space after the start of the label field. If a line of code has no label on it, then your assembly mnemonics may begin two spaces after the line number. At least one space after the op code field will begin the operand field. This field is optional since not all assembly mnemonics have an opcode (such as DEX, or INY). Anything after the operand field is ignored by the assembler, and assumed to be the comment field. A comment can take up an entire line, when the label field begins with a semicolon. Below is a sample line of code with all four fields: LABEL OP OPERAND COMMENT CODE --|-----|-----|-------|------------------------------ 10 START LDA #155 ;Get Carriage return character IF Directive The IF directive is used for "conditional assembly". This may be used to enable, or disable the generation of some "test code", for example, based on the value of a number, label, or expression. For example: 10 DEBUG = 0 ;Enable my debugging test code ... 300 .IF DEBUG @ENDOFDEBUG 310 ; Debugging test routines start here ... 500 ENDOFDEBUG ; End of debugging code If the expression (DEBUG above) is equal to zero, then everything from the line following the IF directive to the specified label (ENDOFDEBUG) is assembled. When you are satisfied that your code works, don't throw away all that useful testing code. Simply change line 10 to DEBUG = 1 and reassemble your program. If you do not understand the use of conditional assembly, don't worry. I have only used it a few times in the past seven years, and generally you don't need it at all. END Directive The Asm/Ed manual recommends that every program have one .END directive, as the last line. It really isn't necessary, since the assembler knows when to stop (it runs out of source code to assemble). If you place a .END in the middle of your program inadvertently, all code after it will be ignored and not be assembled. I seldom use a .END in any of my assembly code. DEBUGGING Your assembly language programs are bound to have bugs in them. Asm/Ed provides a method for testing assembled object code. When at the Asm/Ed EDIT prompt, type BUG and press return. The next prompt will be DEBUG. The commands for the debugger are all short one or two letter commands, some followed by an optional hexadecimal address. To exit DEBUG type X and press return. DR The DR command is used to display the contents of the 6502 registers: DR A=BB X=10 Y=20 P=B0 S=DF A is the accumulator, X and Y are the index registers, P is the processor status register (which includes the carry flag, zero flag, and etc.), and S is the stack pointer. CR The CR command is used to change the contents of any of the 6502 registers. CR<,1,2,,DE The values specified go into the registers in the same order they are displayed by the DR command. In the above example the accumulator is unchanged, the X register receives a 1, the Y register a 2, the status register remains unchanged, and the stack pointer is adjusted to DE. D The D command is used to display memory. D3000,0 tells the debugger to display memory location 3000 hexadecimal. When the second parameter is less than or equal to the first, only one location is shown. D3000,3010 requests the debugger to display memory from locations 3000 through 3010. Enter D by itself and the next 8 locations (3011 through 3018) will be displayed. If only the second parameter is omitted, a default of 8 memory locations are displayed: D3000 3000 10 40 20 22 34 11 12 FE Note that the output of the debugger is always in hexadecimal. All input addresses and register values are to be specified in hex as well. C The C command is used to change memory. The format is: C3034<21,23,,2E The command is immediately followed by the starting hexadecimal address to change. The values to be placed in memory, starting at the first location, are separated by commas. Two commas in a row tell the debugger to skip over that memory location, leaving it unchanged. In the above example memory location 3034 receives 21, 3035 receives 23, and 3037 gets 2E. You may use the D command to display memory just changed, to verify the new values. M The memory move command, M, is used to copy a block of memory from one area to another. The command format is: Mmmmm65535 THEN ? "OUT OF RANGE":GOTO 80 100 PRINT "INPUT NUMBER 2 ";:INPUT N2 110 IF N2<0 OR N2>65535-N1 THEN ? "OUT OF RANGE":GOTO 100 120 SUM = USR( 1536, N1, N2 ) 130 PRINT "NUMBER ";N1;" PLUS ";N2;" EQUALS ";SUM 140 END 1000 PRINT "COULD NOT FIND USR ROUTINE FILE" 1010 PRINT "MYUSR.OBJ" 1020 END Now we need to write an assembly language program with Asm/Ed that implements this USR routine. It will accept parameters N1, and N2 off the stack (two, two byte integers), add them, and return the result to SUM through memory locations $D4 and $D5. Our code might appear as follows: 0 ;LIST#D:MYUSR.ASM 10 ;ASM ,,#D:MYUSR.OBJ 11 SUM = $D4 12 NUM1 = $E0 13 NUM2 = $E2 20 *=1536 ; Assemble for PAGE 6 30 ADDTHEM PLA ; First off the stack is parameter count 40 BEQ ERROR ; Always check for no parameters ERROR 50 CMP #2 ; Did we get exactly 2 parameters? 60 BEQ AOK 70 TAX ; No, clean up stack and return safely 80 CLEANUP PLA ; Two bytes per parameter 90 PLA 100 DEX ; Get all the parameters off? 110 BNE CLEANUP ; when all gone, just the valid return addr 120 ERROR RTS ; is at the top of the stack for the RTS 130 ; We have valid input, compute the sum 140 ; The first parameter in the USR call (after the addr) 150 ; is the first parameter off the stack, high byte 160 ; low byte sequence. REMEMBER this! 170 AOK PLA ; Get NUM1, high byte 180 STA NUM1+1 190 PLA ; Get NUM1, low byte 200 STA NUM1 210 PLA ; Get NUM2, high byte 220 STA NUM2+1 230 PLA ; Get NUM2, low byte 240 STA NUM2 250 ; Now we have the data in temporary storage 260 ; and the stack is cleared of parameters. 270 ; Just the return address (to get us back to BASIC) 280 ; is at the top of the stack - which gets pulled off 290 ; into the program counter automatically by the RTS 300 ; instruction. 310 CLC ; Must clear the carry flag first 320 LDA NUM1 ; Low byte of first integer to add 330 ADC NUM2 ; Add to low byte of second integer 340 STA SUM ; And store in low byte of their SUM 350 LDA NUM1+1 ; Now add high bytes, leave carry alone 360 ADC NUM2+1 ; It "carries over" from previous add 370 STA SUM+1 ; And their summation is complete 380 RTS ; Back to BASIC Enter this program with Asm/Ed and execute the instructions in the first two comment lines. When you get an assembly with no errors, your file D:MYUSR.OBJ should be ready to test with the first BASIC program. Work at this until it performs as expected. As you become more adept at writing USR routines, you may wish to develop utilities for converting OBJ files into a series of BASIC DATA statements, so you can simply READ and POKE them without using messy file I/O to initialize the USR routine. It takes a relatively long time to install USR routines by poking them into memory or strings, but once in place they execute amazingly fast. You will find that USR routines are incredibly difficult to debug. You need to initialize them and call them from BASIC. If you mess up the stack or some other operation, the computer usually crashes inexplicably. It isn't easy to debug USR routines from DEBUG, because you will have to write sophisticated test routines to stuff all sorts of test values on the stack. Stand Alone Assembly Sooner or later you will get tired of USR routines (mostly because they are so difficult to debug). When you do, it is time to take the plunge into writing a stand alone assembly language program. Then you will get into the complexities of keyboard input, screen output, disk I/O, and printer output from the Asm/Ed environment. Complete libraries of routines, such as a "graphics package" that performs the equivalent of BASIC's GRAPHICS, COLOR, PLOT, and DRAWTO, will become a necessity. This is where BOOT CAMP will help the most. In the months to come you will learn everything from keyboard input to floating point processing, all from the assembly language level. Most of our listings are in Mac/65 format. With the exception of macros (Asm/Ed is not a macro assembler), most changes to Asm/Ed compatibility will be minor. As an example of a stand alone assembly language program, and an illustration of its raw speed, we present the following demonstration. First type this BASIC program and run it. While it executes (it will take about 12 minutes), read the remainder of this article to see how the same functions can be performed in assembly language: 10 DINDEX=88:REM Screen RAM pointer 20 SCREEN=PEEK(DINDEX)+256*PEEK(DINDEX+1) 30 FOR X=0 TO 255 40 A=X 50 FOR Y=0 TO 255 60 POKE SCREEN+Y,A 70 NEXT Y 80 NEXT X At location DINDEX is a two byte "pointer". Memory locations 88 and 89 hold the address of the beginning of screen RAM. The equation in line 20 calculates the variable SCREEN, which we use as a direct pointer, for the POKE in line 60. In our assembly language equivalent of the above program, this problem is even EASIER to solve. (This is seldom the case however, most things are much harder to do in assembly language. This demonstration is designed purposefully to show the strengths and speed of assembly language.) Next two loops are setup. The inner Y loop is used to poke the current value of X into the first 256 screen RAM locations. You will see these characters fill the top portion of your display. All ATASCII values from 0 through 255 are poked, with the help of the X loop. The variable A was used simply for a more symmetrical comparison with the assembly code to follow. Let this BASIC program run to completion. Time it carefully, study the sweep second hand of your watch creep slowly along. Feel the annoying impatience of this terribly slow program creep up your spine. When you finally get the READY prompt, reboot your computer with Asm/Ed and enter this equivalent assembly language program: 0 ;LIST#D:SCREEN.ASM 1 ;ASM,,#D:SCREEN.OBJ 2 *=$3400 3 RUNAD=$2E0 10 DINDEX = 88 ; Screen RAM pointer 20 ; We don't have to compute SCREEN, we use post indexed addressing 30 START LDX #0 ; Initialize variables for loops 40 LDY #0 50 STORE TXA ; Place screen character into A register 60 PUTIT STA (DINDEX),Y ; Place character on screen 70 INY ; Next screen location 80 BNE PUTIT ; Y register "wraps around" to zero after 255 90 INX 100 BNE STORE ; NEXT X 110 RTS ; Return control to DOS 120 *= RUNAD 130 .WORD START ; So we can load and run from DOS Now execute the two commands in the first two comment lines at the top of the listing. If you get no assembly errors then you will have a file SCREEN.OBJ that is ready to load and run. Go to DOS and execute a binary 'L'oad of the file SCREEN.OBJ. It will run immediately after loading and return control back to DOS after performing all 65,536 "POKES" of characters to screen memory. Did you catch it? You probably didn't if you blinked. This version of the program takes barely a second to run! If you want to watch the show for a while, and exit to DOS when a KEY is pressed, for example, modify your program as follows: 15 CH = 764 ; Keyboard buffer ... 101 LDA #255 102 CMP CH ; key pressed? 103 BEQ START ; Nope, loop 104 STA CH ; Yes, clear out key buffer and exit to DOS List this version to disk and reassemble it. When loaded from DOS, it will "poke" all those ATASCII patterns to the screen continuously until you press any key on the keyboard. To RANDOMIZE the show, make these changes: 16 RANDOM = 53770 ; Always a random number here ... 50 STORE 60 PUTIT LDA RANDOM ; Get a random fill character 61 STA (DINDEX),Y ; Place character on screen Notice how I always added a meaningful label for each important memory location. Avoid the use of code such as LDA 53770. The proper use of labels makes it much easier to see exactly WHAT your program does and HOW it gets the job done. If you didn't pay much attention to ANALOG's Master Memory Map series, I strongly recommend that you go back and read it all. Even if you do not understand all of it, you will learn a lot. A good memory map is the key to unleashing all the power of your computer. Compute!'s Mapping The Atari, Revised Edition is also a very good reference guide. As a 6502 Assembly language reference manual, I use 6502 Assembly Language Programming by Leventhal. This is a general reference for the 6502 microprocessor, and does not have any specifics on the Atari computer. It does detail all the 6502 assembly mnemonics, and provides examples of multiply, divide, and other useful routines. When you find that Asm/Ed is too slow to suit your tastes, as you build larger and more sophisticated programs, consider upgrading to Mac/65. This macro assembler supports the use of INCLUDE files, allowing you to easily import "canned routines" that have already been debugged. It's MACRO capabilities allow you to define high level constructs that vastly simplify the development of assembly programs. With a good MACRO library (such as OSS/ICD's Mac/65 Toolkit or QuickCode from Stardust Software), your assembly source code will resemble BASIC or some other high level language, while retaining all the power and speed of pure assembly language. Mac/65 is the absolute FASTEST native 6502 assembler I have ever used, bar none. (Mad Mac for the Atari ST will assemble 6502 code that blows the doors off Mac/65; but that's a whole new ball game.) Welcome to the fast and complicated world of assembly language programming. I hope this guide will inspire you to put that inexpensive Asm/Ed cartridge to work on all those fantastic ideas that old faithful Atari BASIC never could handle. By: Matthew J. W. Ratcliff, Ratware Softworks, 32 S. Hartnett Ave., St. Louis, MO 63135